Msg#7862   To:JAMES MOORE                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/01/89
Subject:(R)CONDITIONS IN .FRM'S                 Parent Msg#7837
------------------------------------------------------------------------------
What does _isonline() do? What is the actual code? I had a somewhat
similar problem, I was doing  COPY TO (filename) FOR &expr WHILE
CountRec() SDF  to make an SDF file for the expression EXPR, and here's
CountRec()
 
FUNCTION CountRec
@ ... SAY RECNO()/RECCOUNT()*100   with a transform to make it xx.x%
RETURN(!EOF())
 
It would get to the COPY... line, and 16.7% would be displayed, and then
it would crash. Why? Because of something in the way Clipper handles
RECNO(), RECCOUNT() or EOF(), while doing, oh, say COPY TO, that bombs
the program.
 
BTW, in the previous message, it looks like "...FOR field1 > field2.
.and. ..." Is there a period (.) after field2 (meaning an extraneous
dot)? Or is this just line noise/typo?
 
What are field1 and field2 ? Are they, say, "FOR TotalOvertime >
HisMaxOT" or something like that? What TYPE() are they, how large (C15,
N18.2, what?)
 
I do not know what isprinter() does internally, but I can make a damn
good guess. INT 15, Subfunction something I believe are the printer
input, printer output and printer status functions. It checks the return
value from the printer status INT, and if certain bits are/are not set,
it returns .T./.F., appropriately. It could be the REPORT FORM command
does not like to have isprinter() invoked from within it, either for a
good reason, or Clipper's fault. I do not know, I never pulled these
functions apart. Perhaps Dirk could shed some light?
 
Sorry if I couldn't be very helpful. Luck...


Msg#7863   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/01/89
Subject:(R)NANTUCKET                            Parent Msg#7695
------------------------------------------------------------------------------
That makes sense, since if Clipper is smart (or MSC, as the case may be)
all registers are either: left to the user to save/restore or
pushed/poped by the compiler before/after each external call. I'd bet on
the former though. Will be playing with TC/Clipper soon, I'll let you
know how things turn out. 


Msg#7865   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/01/89
Subject:(R)WP MERGE                             Parent Msg#7792
------------------------------------------------------------------------------
Is that:
 
FIRST^RSECOND^RLAST^R^ENEXT_FIRST^RNEXT_SECOND^RNEXT_LAST^R^E
 
or
 
FIRST^R^JSECOND^R^J...^E
 
where each record + ^R is followed by a LF (^J)? Also, if a ^J is needed,
does it go: 1) between the field and the ^R, or 2) after the ^R? Also,
does there have to be a ^J between the ^R and ^E at the end of a record?
 
Thanks. Please reply as soon as possible, this is one of the two last
functions awaiting to be completed. The other is just a report that I
need to find the time to sit down and write.


Msg#7866   To:JON MELTZER                               (Rcvd)
From:DIRK LESKO                                 Date:02/01/89
Subject:(R)LINT FOR CLIPPER                     Parent Msg#7841
------------------------------------------------------------------------------
Another product? OK, who wants to write it, and will they be able to beat
Nantuckets next release?
 
dLESKO


Msg#7867   To:STEVE DAVIES                               
From:DIRK LESKO                                 Date:02/01/89
Subject:(R)CONDITIONS IN .FRM'S                 Parent Msg#7837
------------------------------------------------------------------------------
If you can't find frm2prg, try contacting Russel Freeland, he wrote it
and it should be on his board...
 
dLESKO


Msg#7868   To:RON LANDBERG                              (Rcvd)
From:DIRK LESKO                                 Date:02/01/89
Subject:(R)A86 TO S87                           Parent Msg#7843
------------------------------------------------------------------------------
I have had some problems with arrays inside valid statements, and any
third party libraries you are using must be checked...
 
dLESKO


Msg#7869   To:CONSTANTIN FLORESCU                       (Rcvd)
From:DIRK LESKO                                 Date:02/01/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Is it only the ALT/E key, or does it not work with other keys as well? If
it bombs on other keys I would change the name of the procedure to not
include the word 'DO' at the beginning, you might be fooling the parser.
If so, then another 'anomoly' to add to the list....let us know..
 
dLESKO


Msg#7870   To:ART KOHN                                  (Rcvd)
From:DIRK LESKO                                 Date:02/01/89
Subject:(R)PROGRAMMER WANTED                    Parent Msg#7850
------------------------------------------------------------------------------
Make me an offer....(just kidding)


Msg#7871   To:ANDREW GROSS                              (Rcvd)
From:DIRK LESKO                                 Date:02/01/89
Subject:(R)ERROR 10                             Parent Msg#7807
------------------------------------------------------------------------------
The reason why stack errors happen at 16 is because that is the max
number of 'levels' you can nest anything, including SET KEY TO stuff...
 
dLESKO


Msg#7872   To:DIRK LESKO                                (Rcvd)
From:DAVID MORGAN                               Date:02/01/89
Subject:(R)NETERR()                             Parent Msg#7719
------------------------------------------------------------------------------
 
As I understand it DOSERROR() is only for use inside the error
system.  The value it returns elsewhere isn't going to tell you
anything meaningful.  The limitation may be greater, i.e., for use
only within open_error().  That's the only one from which it's called
in the default error funtions.  It's a limited purpose function and
from that standpoint the generality of the name is unfortunate.


Msg#7873   To:DIRK LESKO                                (Rcvd)
From:DAVID MORGAN                               Date:02/01/89
Subject:(R)EMS FOR THE PC                       Parent Msg#7753
------------------------------------------------------------------------------
Dirk, David Schwartz, Howard Kapustein--
 
>> "Discord" between Clipper and certain EMS drivers
 
I've never used an AST Rampage but a number of people that use it
with Clipper have reported various problems, right from the beginning
with Autumn '86 when Clipper first made use of EMS.  I can't tell you
what concretely.  They often say that their problems go away when
they SET CLIPPER=E0 to disable use of EMS.  Clipper makes use of EMS
by issuing the calls for that purpose specified by (L)otus, (I)ntel,
and (M)icrosoft in their LIM specification.  Intel will (or used to)
send you a copy (about 30 stapled pages) on request.  Every EMS
hardware manufacturer and his brother supplies his own software
driver to implement the responses to the various LIM calls, no doubt
all different internally.  AST for its part follows a standard that's
explicitly different from LIM (AQA - Ashton-Tate, Quadram, AST).
It's supposed to be a superset of LIM, and
LIM-application-transparent.  But maybe the incidence of problems
with AST's particular driver has to do with the AQA-vs-LIM
differences.


Msg#7874   To:TERRY DILLARD                             (Rcvd)
From:DAVID MORGAN                               Date:02/01/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
 
Make sure the dependency lines in your MAKE file have no trailing,
invisible characters (even spaces) after the name of the last file on
the line.  There should be an *immediate* CR/LF (I think) or it
confuses the intervening invisible characters as part of the file
name -- which named file "does not exist."  Other MAKEs are more
robust than Nantucket's, and MAKEs are not product specific.  If you
have other language compiler products you may well have another MAKE
on your shelf.  A lot of people seem to like Borland's.


Msg#7875   To:MARK NEIDORFF                             (Rcvd)
From:RON LANDBERG                               Date:02/01/89
Subject:(R)DBU BUG                              Parent Msg#7784
------------------------------------------------------------------------------
Also occurs in changing field from or to logical.  It's just an annoying
bug and you have to leave DBU and then reenter it to get around the
problem.


Msg#7876   To:ROB HANNAH                                (Rcvd)
From:RON LANDBERG                               Date:02/01/89
Subject:(R)NOVELL 2.0A LIMIT                    Parent Msg#7759
------------------------------------------------------------------------------
On Rob's advice, I used this patch many months ago.  It works great.


Msg#7877   To:RON LANDBERG                              (Rcvd)
From:LARRY GOTTLIEB                             Date:02/01/89
Subject:(R)A86 TO S87                           Parent Msg#7843
------------------------------------------------------------------------------
I ran across a problem with CONTINUE.  It seems that A86 is more
forgiving in this area.  The string of code went something like this:
 
       USE myfile
       *  Note that no LOCATE command has been issued.
       CONTINUE
 
This is clearly incorrect.  However, A86 seemed to treat this as a SKIP;
S87 positions the record pointer to end of file.


Msg#7880   To:ANDREW GROSS                              (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/01/89
Subject:(R)WORKHORSE PRINTER                    Parent Msg#7644
------------------------------------------------------------------------------
Thanks Andrew... I will take a look at it.
<CF>


Msg#7882   To:DIRK LESKO                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/01/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Dirk, doesn't even get to the procedure like I told John Booth yerstaday.
The actual name of the procedure is "Expedit" but doesn't work regardless
of what name it is. Mark(Neidorf) just said that probably I can use the
keys up to value of 255. But then that means that I can't use a
combination of ALT+ alpha character ? I think that should be possible!
Thanks Dirk. BTW Do you think I should be able to SET KEY TO "ALT+ alpha
character"?
<CF>


Msg#7883   To:TERRY DILLARD                             (Rcvd)
From:DAVID MINTER                               Date:02/01/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
You seem to have the statement DEBUG in the wrong place.  If you are
trying to link in clipper's debug.obj, be sure that it is after your
files command and not somewhere else.
 
Hope I can get an ATTABOY.
 
Dave


Msg#7884   To:CONSTANTIN FLORESCU                       (Rcvd)
From:JOE BOOTH                                  Date:02/01/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Try to isolate the problem.  First, try assigning the pop-up to -1 (F2)
or some key.  See if that works.  Then try writing a small program with
an INKEY() statement, and see what number is being generated when you 
pressed ALT-E.  Possibly the BIOS is intrepeting the scan code
differently than the manual.   Good luck.


Msg#7885   To:FRANCOIS DESROSIERS                        
From:JOE BOOTH                                  Date:02/01/89
Subject:(R)MEMOEDIT                             Parent Msg#7861
------------------------------------------------------------------------------
A quick and dirty solution would be to use a character field, say 4000
characters long, instead of a memo.  Memoedit can work on either a
character field or a memo field.  The obvious drawbacks are that you 
might not have enough character space reserved in the case of a long
memo, or you might have a lot of wasted space.  Good luck


Msg#7888   To:ALL                                        
From:NEIL WEICHER                               Date:02/01/89
Subject:GET-IT FUNCTION LIST     
------------------------------------------------------------------------------
In response to inquiries, I am posting the full list of Get-It functions.
 
 N_ANYKEY     Calls procedure when any key is pressed.
 N_DELGET     Remove Get from input list.
 N_GETALIAS   Returns alias of specified input field.
 N_GETCOL     Returns column position of specified input field.
 N_GETLEN     Returns display length of input field.
 N_GETNO      Returns input field number.
 N_GETPIC     Get or set picture associated with specified input field.
 N_GETROW     Returns screen row position of specified input field.
 N_GETVAR     Returns name associated with specified input field number.
 N_GOTOGET    Move cursor to specified input field.
 N_LASTGET    Returns number of input fields on screen.
 N_MOVEGET    Dynamically changes order of input fields fields.
 N_NEWFLD     Calls procedure when cursor moves to new input field.
 N_NOKEY      Calls procedure if no key pressed within specified time.
 N_ONCOMx     Calls procedure when carrier or ring detected at COMx.
 N_ONSHIFT    Calls procedure when shift or lock is pressed or released.
 N_ONEXPR     Calls procedure when <expression> is true.
 N_ONTICK     Calls procedure repeatedly on a regular interval.
 N_REGET      Redisplay input fields (called after N_RESTGETS).
 N_RESTGETS   Restore gets.
 N_RESTSCR    Restore screen attributes.
 N_SAVEGETS   Saves gets.
 N_SAVESCR    Saves screen attributes.
 N_SCHAR      Returns character(s) at screen location.
 N_SCOLOR     Returns color clause at screen location.
 N_SPLITA     Returns alias portion of specified `alias->name' clause.
 N_SPLITN     Returns name portion of specified `alias->name' clause.
 N_WSTATE     Returns wait-state type.
 
For more information on Get-It, call 212-724-0150 or write: 
Communication Horizons - 701 7th Ave. #900 - New York, NY - 10036.
 
Note: current owners may update at no charge by mailing back the 
original disk in a STAMPED, SELF ADDRESSED diskette mailer.
 


Msg#7889   To:CLIFF GREEN                               (Rcvd)
From:STEVE DAVIES                               Date:02/01/89
Subject:(R)CONDITIONS IN .FRM'S                 Parent Msg#7837
------------------------------------------------------------------------------
It's right here on The Boss
Steve


Msg#7893   To:DAVID MORGAN                               
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)NETERR()                             Parent Msg#7719
------------------------------------------------------------------------------
I have not hacked DOSERROR to death, but I do know that MSC keeps an
error vairable called __oserr. Do you know, or can you find out if it
does indeed contain what DOSERROR returns. WOuld save me some trouble
looking for it....
 
dLESKO


Msg#7894   To:DAVID MORGAN                               
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
There is a file on the board for making called NDMAKE. ANd it's pretty
good. Got lots of looping and logic control as well....
 
dLESKO


Msg#7895   To:CONSTANTIN FLORESCU                       (Rcvd)
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
I do it all the time. ALT+many keys works great. Are you sure that
inkey() actually returns the value in question? some keyboard bios do not
handle certain ALT key combinations (just try ALT+UP ARROW) to name
one....
 
dLESKO


Msg#7897   To:DAVID MORGAN                               
From:HOWARD KAPUSTEIN                           Date:02/02/89
Subject:(R)EMS FOR THE PC                       Parent Msg#7753
------------------------------------------------------------------------------
LIM = Lotus-Intel-Microsoft = EMS 
AQA = AST-Quadram-AshtonTate = EEMS
 
EMS allows bank switching, and EEMS was merely a superior form of EMS,
allowing a 64K swapped bank to relate to, not one 64K block, but 4 16K
blocks. I'm pretty sure that was the definition, but if not, it's close.
Anyway, EEMS is EMS + a little more.
 
Intel will send you, free, the LIM EMS specs (now 4.0) for free. Sorry, I
don't remember the number.
 
As of LIM EMS 4.0, EMS and EEMS are FUNCTIONALLY the same. However, EEMS
is slightly superior due to the way the hardware works. BTW, try to find
a memory card that supports 4.0 by hardware, not software. I think the
AST cards do, but don't quote me on that.
 
Oops, re: LIM EMS specs, they are on 5" x 8" ? The same size as the IBM
manuals. No binder, just small pages already hole punched.
 
My question still remains: Who sells an EMS card for the PC? Virtually
every memory card available is for the AT on up. AST makes Rampage/2, the
now discontinued 6PakPremium, and STB has their Memory Companion/PC. No
one else knows of any other manufacturer. I have talked with over a dozen
sales reps and technical people from mail order companies, 47th St.
Photo, etc. and if they even know what I'm talking about, they've only
heard of AST. Maybe. But I haven't heard from anyone who has some actual
hands on experience with an EMS card for the PC.
 
Sidebar: EEMS "should" be better than EMS, all other factors being equal
(same speed RAM, etc.) more efficient, what have you. Is this really the
case? How much is EEMS worth over EMS?


Msg#7898   To:JOE BOOTH                                 (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/02/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Thanks Joe but you wouldn't beleive what I was doing !
As a matter of fact I couldn't sleep thinking about it and all of a
sudden I remember that...before I used these keys(ALT+E etc) in another
procedure and it worked. This morning first thing I did... is take a look
to see how I did then...and found out that insted of
    SET KEY 274 TO Expedit
I was saying
    SET KEY -274 TO Expedit
I was carried over with this "-" from the function keys...
Thanks very much again...Speak to you soon
<CF>


Msg#7900   To:DIRK LESKO                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/02/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Over night I remember I did it before, then this morning I just realized
that I was saying:
    SET KEY -274 TO Expedit
instead of 
    SET KEY 274 TO Expedit
See it...
Sorryy...
BTW Are you back home ? Do you remember by any chance the name of the
person you said you would look up?
Thanks. Speak to you...
<CF>


Msg#7901   To:ART KOHN                                  (Rcvd)
From:STEVE BADARACCO                            Date:02/02/89
Subject:(R)PROGRAMMER WANTED                    Parent Msg#7850
------------------------------------------------------------------------------
About the position ...
I live in Eastern Connecticut and work in Hartford right now.
I do have experience with Clipper and Funcky; I've been doing almost
exclusively Clipper work for over 2 years, and I'm VERY good at it.
 
If you can waive any requirements for travel to NYC, then I might be
interested.  I don't mind doing work long-distance, I currently support a
couple of dozen field offices, some running LARGE Clipper programs which
I have written.
 
Let me know; if this is acceptable, we can pursue further.
 
... Steve Badaracco


Msg#7902   To:HOWARD KAPUSTEIN                          (Rcvd)
From:MARC SCHNAPP                               Date:02/02/89
Subject:(R)EMS FOR THE PC                       Parent Msg#7753
------------------------------------------------------------------------------
Howard, EEMS is a mot point 'cause EMS 4.0 was an incorporation of the
AQA EEMS standard.  The current Intel products ARE EEMS compatible.
   
By the way-- I have used Rampage boards almost exclusively and have NEVER
had any problems-- that includes client sites.  You mention a problem
with AST and Desqview--- that's incoceivable 'cause they were bundled
(may still be).  I never had any such problem runing the AST oem version
or the comercial version of Dersqview.
  
Your problem is obtaining an 8-bit card.  I would avoid STB just on gut
feel-- who will be there to support the card in a couple years when STB
gets out (they probably will)..    I would probably get Intel if I had a
choice-- they are very comitted (for now).  Quadram also had an 8-bit
card for a while-- the QuadEMS I/O which does EEMS and EMS.  It probably
is discontinued but call Quadram 404 923-6666.  I have one with 1 meg--
but I think I already have a customer.  I'll know in 10 days-- if not,
you're welcome to it.   
   
If you settle down to a new one-- let me know-- I can sell hardware (I am
a lowkey dealer) and I will do my best (esp. since you've been doing YOUR
best for me).
   
Finally re: discord with Rampage and Clipper (re: Dave Morgan):  I
suspect that the problem may be sensitive to the version of the
driver'cause Pat Adams complained only after changing drivers.   I am
hapy with the Rampage cards.  I have several 2meg boards (used) to sell
now that I have a 386, but they have served me and my clients faithfully.
  
One more thought-- there is a CA liquidator that periodically offers
Quadram products-- they usually advertise in the back of Infoworld.


Msg#7905   To:ALL                                        
From:MARC SCHNAPP                               Date:02/02/89
Subject:RAMPAGE                  
------------------------------------------------------------------------------
If you haven't been scared away from Rampage cards-- I have several
Rampage 16-bit cards that I cam no longer using since I acquired a 386.
Each has 2megs (a gold mine right there).
   
Please let me know if you are interested in getting some.


Msg#7906   To:DIRK LESKO                                (Rcvd)
From:ART KOHN                                   Date:02/02/89
Subject:(R)PROGRAMMER WANTED                    Parent Msg#7850
------------------------------------------------------------------------------
i learned a long time ago not to go after the unatainable... 


Msg#7908   To:DIRK LESKO                                (Rcvd)
From:TERRY DILLARD                              Date:02/02/89
Subject:(R)CLIPPER HELP          
------------------------------------------------------------------------------
Dirk,
Thanks for the help, buI'm pretty sure that I've got my MAKEFILE the way
the manual (and you) say it should be.  I can delete all .OBJ files and
run MAKE <makefile> and it compiles and links perfectly.  I only get the
"dependant" message when there are EXISTING .OBJ files.  As for the
overlay mess, I'll look into it, you could have solved it for me.  Thanks
for the help.


Msg#7909   To:JOE BOOTH                                 (Rcvd)
From:TERRY DILLARD                              Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
Thanks, Joe.  I;ll check out the spelling on the .EXE line, although I'm
pretty sure they're all OK.  As far as the DEBUG statement goes, I'm
using Borland's TLINK, and don't have one, I don't think.


Msg#7910   To:DAVID MORGAN                               
From:TERRY DILLARD                              Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
I'll check on that, thanks!


Msg#7911   To:DAVID MINTER                              (Rcvd)
From:TERRY DILLARD                              Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
I don't believe I have a DEBUG statement anywhere in the MAKEFILE or the
LINK file.  I'll go back and check, but I KNOW I'm not linking in the
Debugger, and I'm pretty sure the only statements in the make file are
the .OBJ : .PRG dependencies, and the .EXE : .OBJ dependencies at the
end, along with the associated commands.  This is an example of what I'm
using, could the problem possibly be the multiple .PRG statements after
UTILMENU.OBJ?:
  
TRACK.OBJ : TRACK.PRG
  clipper TRACK -m
OPENSCRN.OBJ : OPENSCRN.PRG
  clipper OPENSCRN -m
UTILMENU.OBJ : UTILMENU.PRG BACKUP.PRG RESTORE.PRG PACK.PRG VIEW.PRG
  clipper UTILMENU
  
UTILMENU calls the other four files, so CLIPPER will compile them into
one .OBJ file.  Could this be the problem.  I don't think so, but I'll
listen to anyone who can help.  Thanks.


Msg#7913   To:TERRY DILLARD                             (Rcvd)
From:DAVID MINTER                               Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
If I remember right, you said that whenever your app loads an overlay,
you got the message "Loading overlay XXXX" or something like that.  This
is a PLINK86 feature, not any make program feature.  If your not using
PLINK86, I'm not exactly sure what is going on.  But if you are, more
than likely you have a DEBUG statement somewhere in your link file. 
(I've used it a few times to try and determing where things hang up).
 
Good Luck,
Dave


Msg#7914   To:TERRY DILLARD                             (Rcvd)
From:JOE BOOTH                                  Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
I thought you send you were making overlays?  TLINK cannot make overlays.


Msg#7915   To:TERRY DILLARD                             (Rcvd)
From:JOE BOOTH                                  Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
Remove the other program calls from the UTILMENU line  and give it a try


Msg#7918   To:ALL                                        
From:NOEL MANZANILLA                            Date:02/02/89
Subject:LASER PRINTER                           Parent Msg#7918
------------------------------------------------------------------------------
help, I am going to do an application using a laser printer and Clipper
(of course).  I seem to remember there are products that allow you to 
define where a string(as well as fonts) will be printed and this can 
be done from whithin a Clipper application.  the trouble is I can't find
the @&*#! add anymore.  would appreciate all the help you can give me.
(eg using @ say's with a laser printer.)  also, I would like to know if 
there is some library out there that would allow me to do forms on the
laser.  thanks for the help.
noel


Msg#7919   To:CONSTANTIN FLORESCU                       (Rcvd)
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Look up who? I have forgotton already....And I am still out here in Chi
town, lots of work...
 
dLESKO


Msg#7921   To:ART KOHN                                  (Rcvd)
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)PROGRAMMER WANTED                    Parent Msg#7850
------------------------------------------------------------------------------
But...you will never know if it is unattainable until you try to go after
it, right?
 
dLESKO


Msg#7922   To:TERRY DILLARD                             (Rcvd)
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
Have you tried seperating your  commands freom the next line of text with
a space? maybe the make file is thinking it is part of the command of the
previous dependancy...
 
dLESKO


Msg#7925   To:NOEL MANZANILLA                           (Rcvd)
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Isn't there a file on this board with a bunch of Laser stuff in it? Take
a loook, I know I've seen one (primitive of course) - the people in the
AD that have disappeard were from Mayflower consulting, they may still
have an add for there little network library, they used to advertise the
HP thing too, but I don't see it anymore...
 
dLESKO


Msg#7926   To:ALL                                        
From:RICHARD HORWITZ                            Date:02/02/89
Subject:BEGIN SEQUENCE                          Parent Msg#7926
------------------------------------------------------------------------------
 I was just wondering....what is the BEGIN SEQUENCE/END structure good
for? I have seen people make referance to it, but I'm not really sure
what I would use it for. Is it like a rollback command in transaction
processing? Any feedback would help, Thanx.


Msg#7927   To:RICHARD HORWITZ                           (Rcvd)
From:STEVE DAVIES                               Date:02/02/89
Subject:(R)BEGIN SEQUENCE                       Parent Msg#7926
------------------------------------------------------------------------------
Well, besides being a little dangerous, it does things like RETURN TO
MASTER, GOTO (ugh), EXITing out of several loops, and I think you can
issue it from errorsys to jump over ssome code like a print routine when
the user can't (or doesn't want to) fix the printer. 
Steve


Msg#7928   To:RICHARD HORWITZ                           (Rcvd)
From:DIRK LESKO                                 Date:02/02/89
Subject:(R)BEGIN SEQUENCE                       Parent Msg#7926
------------------------------------------------------------------------------
think of begin sequence of a way of doing a goto without having the
'real' programmers scoul on you for using it. Particulary it is good for
simulating the RETURN TO MASTER command when you want something like an
error routine to return to the main menu on errors, however I prefer to
avoid errors in the first place....
 
dLESKO


Msg#7929   To:ALL                                        
From:DIRK LESKO                                 Date:02/02/89
Subject:WORDPERFECT 5.0                         Parent Msg#7929
------------------------------------------------------------------------------
     
Since I have not seen the definitive answer to the WordPerfect 5.0
Secondary Merge file format, I thought I'd share it with you. After
updating my Mail-Merge File Manager which now supports 10 types
of file types, I have discovered the following attributes of
WordPerfect 5.0.
  
It does not use the exact same format as 4.x, although they appeared
to be similar at first glance. WordPerfect 5.0 also introduces
another merge file format which they call a database file {chuckle}.
The database file is created in WordPerfect 5.0 and it uses tabs 
to seperate each field, and a newline to seperate each record. The
'database' can then be sorted, and IF OR AND operations can be
performed on it. Once it is sorted it is converted into a secondary
file for merging. This made it a little easier for novices to
create secondary files. The format for this type of database is:
     
        <field1> chr(9) <field2> chr(9) <field3> chr(9) chr(13)+chr(10)
        <field1> chr(9) <field2> chr(9) <field3> chr(9) chr(13)+chr(10)
        <field1> chr(9) <field2> chr(9) <field3> chr(9) chr(13)+chr(10)
        <field1> chr(9) <field2> chr(9) <field3> chr(9) chr(13)+chr(10)
  
Where each line is a record that ends in a hard carriage return. The
text of each field can contain spaces, but they must be seperated
by tabs (chr(9)). tab stops must be set so that proper formatting
of the fields take place if they want it to look good on screen -
(meaning all the tabs line up).
   
The secondary file is different in a tiny, but annoying way. the
layout of a secondary merge file is as follows, where each block
represents a record.
   
        <field1> chr(18)+chr(13)+chr(10)+;
        <field2> chr(18)+chr(13)+chr(10)+;
        <field3> chr(18)+chr(13)+chr(10)+; 
        <field4> chr(18)+chr(13)+chr(10)+; 
        <lastfield> chr(5)+chr(12)
    
        <field1> chr(18)+chr(13)+chr(10)+;
        <field2> chr(18)+chr(13)+chr(10)+;
        <field3> chr(18)+chr(13)+chr(10)+; 
        <field4> chr(18)+chr(13)+chr(10)+; 
        <lastfield> chr(5)+chr(12)
  
The difference is that each record now ends in a chr(5) PLUS a hard
page break character (chr(12)), and without a newline character. That
is the only difference between 4.x and 5.x that I can see.
  
Hope all of you who have been looking for this can make use of it.
  
Yours Truly,
   
 Dirk Lesko
 


Msg#7930   To:ANDREW GROSS                               
From:CLIFF GREEN                                Date:02/02/89
Subject:(R)MLCOUNT                              Parent Msg#7781
------------------------------------------------------------------------------
yeah, I've scanned that article ("The Power of the Memo Field"), and want
to play with it some ... it would help in a project I shelved some time
ago.


Msg#7931   To:HOWARD KAPUSTEIN                          (Rcvd)
From:CLIFF GREEN                                Date:02/02/89
Subject:(R)WP MERGE                             Parent Msg#7792
------------------------------------------------------------------------------
lemmesee.
While _in_ WP, it looks like each ^R and ^E is followed by a CRLF pair (I
honestly haven't looked at the hex representation, so can't tell if WP is
just interpreting ^R or ^E so mere humans can read records nicely).  On
the other hand, I've gone back and removed the CRLF between an ultimate
^R and its ^E (so the line segment looked like:  ^R^E) and had no
difficulty.
Offhand, I'd have to say that you want to put a CRLF between each ^R or
^E and the _following_ data;  if you were to put CRLF, CR, or LF _before_
the ^R, for example, it would show up in your merge product (not usually
a pretty sight, except if you're merging multi-line items - like the
return address at the top of a business letter).


Msg#7932   To:ANDREW GROSS                               
From:TIM CREAGH                                 Date:02/02/89
Subject:(R)ERROR 10                             Parent Msg#7807
------------------------------------------------------------------------------
Andrew, thank you for responding to my Error 10.  I was beginning to
think no one knew what it was.  I am using the Sequence command and it is
very possible that I over looked a LOOP inside one.  I'll go in tonight
and check.
Thanks again, that'll sure save alot of agravation.
-tim-


Msg#7933   To:NOEL MANZANILLA                           (Rcvd)
From:CLIFF GREEN                                Date:02/02/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
If you're going to be moving the cursor around on the page in a laser
printer, you're much better off writing some UDF's which include the PCL
code for moving the cursor, and then printing them with SET PRINT ON and
?? instead of @SAYs.  That way, you don't increment any line counters that
Clipper may be keeping.
I just finished a project which depended on controlling a laserjet, and
found Clipper and PCL (plus a few font files) to be more than adequate. 
This was one of those satisfying projects:  the State of New Jersey is
going to take my printer output and publish it untouched.


Msg#7934   To:NOEL MANZANILLA                           (Rcvd)
From:CLIFF GREEN                                Date:02/02/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
by the way, what do you need to do on the laser printer, and *what brand*
is it?  (e.g., I've been working with HP LaserJets mostly, but have
noodled around with Kyoceras, and the same principles should apply,
though the printer control codes are very different).


Msg#7936   To:MARC SCHNAPP                               
From:HOWARD KAPUSTEIN                           Date:02/03/89
Subject:(R)EMS FOR THE PC                       Parent Msg#7753
------------------------------------------------------------------------------
Thanks. Yes, EMS is functionally equivalent to EEMS as of 4.0, but isn't
it more efficient to have that support via hardware (EEMS) than software
(EMS). Many of the cards out there now are EMS, and those that are now
4.0 compatible do it via new software (mainly the EEMS cards) or people
like Intel, etc. the industry leaders. Although with their prices it'll
be quite a while until I get one of those. 
 
I know AST supports 4.0, but I believe it's through a new driver, no
hardware mods. I'm not sure who has a 4.0 hardware supported card. 
 
STB is a semi-reliable name. Besides, I'm not too concerned with them
being around in 4 years, if the card is up to snuff, there should be no
problems, and I plan on getting a newer machine in much less than 4
years.
 
I just found out the AST problem (I think): with certain 4.x drivers, and
communications under DV, because of the switching done by the EMS driver
with background communications (generally high speed) the machine can
potentially lock up. I believe it's an exponential curve up, the higher
the baud rate. 
 
If I can get enough solid information regarding the AST 6PakPremium, and
the price is right, I'll get it. The stumbling block is finding a good
price for one with RAM, and verifying that it is 4.0 compatible. I'm
still suspicious that they discontinued it, since it was reasonably
popular among EMS 808x users, and their Rampage/2 is double the price for
the same amount of RAM, but minus the clock, serial and parallel ports.
The IBM marketing apporach: Offer less for more...


Msg#7937   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/03/89
Subject:(R)WORDPERFECT 5.0                      Parent Msg#7929
------------------------------------------------------------------------------
When you say the sec. format is "...different in a tiny, but annoying
way." I presume you mean the 5.0 sec. format is different from the 4.x
format. Correct?
 
Thanks for all the info, now I can finally finish this program...


Msg#7938   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/03/89
Subject:(R)WP MERGE                             Parent Msg#7792
------------------------------------------------------------------------------
OK, that's what I thought, but I wanted to make sure. Thanks a lot.
Between you and Dirk I have enough info to finish up a program I'm
working on <groan>. Gnite...


Msg#7941   To:ALL                                        
From:STEVE BADARACCO                            Date:02/03/89
Subject:PLINK86 FOR SUMMER 87                   Parent Msg#7941
------------------------------------------------------------------------------
I have been linking some HUGE applications lately, all of which include
the libraries CLIPPER, EXTEND, OVERLAY, FUNCky, and one of my own.
 
I have had a consistent problem with the applications hanging or doing
really bizarre things with the screen and with memory.  This only happens
when I link the overlayed applications, and do not put the line "OVERLAY
PROG,$CONSTANTS" in my LNK file.
Now, from what I understand, the correct way to say this for S87 is
"OVERLAY CODE, $CONSTANTS".  No luck.
And if I leave the line out completely, no luck either!
 
The reported load size, with the "OVERLAY CODE,$CONSTANTS" is about 30K
SMALLER than the load size with the "PROG" version, but it simply will
not run without completely hanging.  Does anyone have a clue as to what
is going on here?
I've been doing nested overlays for a couple years, since the Autumn '86
version, with no problem.  Everything is running on IBM PS/2 Model 60,
very standard hardware all around, no Rampage!.
 
...Steve


Msg#7942   To:STEVE BADARACCO                           (Rcvd)
From:DAVID MINTER                               Date:02/03/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Steve,
Instead of having OVERLAY CODE,$CONSTANTS in your link file, just use
OVERLAY CODE.  
 
Dave


Msg#7943   To:DIRK LESKO                                (Rcvd)
From:ROB HANNAH                                 Date:02/03/89
Subject:(R)BEGIN SEQUENCE                       Parent Msg#7926
------------------------------------------------------------------------------
I use begin sequence often in report routines.  It allows me to quickly
bop out in the middle of a report without resorting to extra conditionals
in the loops.  Here's what I mean:
 
procedure old
private abort
 
abort = .f.
do while ! eof() .and. ! abort
  do while nextkey() > 0 .and. ! abort
    abort = inkey() = 27
  enddo
  if ! abort
    * print stuff
  endif
enddo
 
...
 
procedure new
begin sequence
do while ! eof()
  do while nextkey() > 0
    if inkey() = 27
      break
    endif
  enddo
  * print stuff
enddo
end
 
This becomes more useful the more nesting levels (do while) you have.
 
ROB


Msg#7944   To:ALL                                        
From:CHRISTOPHER EDGAR                          Date:02/03/89
Subject:RANDOM RECORDS                          Parent Msg#7944
------------------------------------------------------------------------------
 
I would like some ideas on how to address a design issue for an
application that will be used to study Foriegn languages.  .DBF files
wil hold nouns and verbs, each record will contain a verb, for example,
along with each of its' respective gender and plural tenses.
 
What I would like to have happen is to be able to have happen is to
"scroll" through the entire file (3000 records) in a random manner,
without repeating any of the records that have all ready come up.  This
should happen like we used to study multiplication tables in a "flash
card" type of scenario.
 
I've played around with the Steve's RANDOMIZE() function to have a random
record be called up, but that doesn't allow for excluding repeats.  Also,
because is tied into the clock, there seems to be a recognizable pattern
to the randomization.
 
Anyone have any Suggestions?
Thanks,
 
--Christopher
 


Msg#7945   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/03/89
Subject:(R)WORDPERFECT 5.0                      Parent Msg#7929
------------------------------------------------------------------------------
Yes, annoying in ythe fact that they really don't tell you the changes
made, only the keystrokes needed to create one in WordPerfect. Also, they
changed the ^p^p convention to a ^s^s convention, and they didn't clarify
the changes....that's all.
 
dLESKO


Msg#7947   To:STEVE BADARACCO                           (Rcvd)
From:DIRK LESKO                                 Date:02/03/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Isn't the OVERLAY CODE,$CONSTANTS thing automatic? (as reported by SJS)
all that tells it is tpo put the CODE segment, and not the Data into the
overlay, so when you say OVERLAY PROG,$CONSTANTS you are probablt ending
up with more data in the overaly, hence the smaller load size. However,
the shared data may cause thrashing which will hang your app..
 
dLESKO


Msg#7948   To:CLIFF GREEN                               (Rcvd)
From:NOEL MANZANILLA                            Date:02/03/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
thanks, can you expand on this .  you are dealing with a first time
laser printer user.  I can't go ahead and ask my company to buy a laser
printer until I'm sure that this will work.  thanks again


Msg#7949   To:CLIFF GREEN                               (Rcvd)
From:NOEL MANZANILLA                            Date:02/03/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
i will be working with HP's unless I find a cheaper alternative.  Any
suggestions ?


Msg#7950   To:CHRISTOPHER EDGAR                         (Rcvd)
From:DIRK LESKO                                 Date:02/03/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Why not index on the key field, and open the file in two work areas.
every time you scroll to a new record in area one, seek the key in area
2, if the record found is before the one your on, you've already seen it,
so skip to the next record in area 1 and repeat....FUNCky's random()
function produces an absolutely random number...
 
dLESKO


Msg#7951   To:JOE BOOTH                                 (Rcvd)
From:TERRY DILLARD                              Date:02/03/89
Subject:(R)CLIPPER HELP                         Parent Msg#7839
------------------------------------------------------------------------------
Sorry about that.  I use TLINK when I don't need overlays, PLINK86 when I
do.  I just recently started using overlays, and haven't adjusted my
vocabulary yet!  You're absolutely right!


Msg#7952   To:ALL                                        
From:TERRY DILLARD                              Date:02/03/89
Subject:ATTABOY!                                Parent Msg#7952
------------------------------------------------------------------------------
To all those who helped since I last read my messages, but especially
to the guys I've already read, your solutions to the overlay problem
worked!  I also just got my MAKE file on the right track, too, it looks
like.  The only thing I can figure is that I had another .MAK file
(TRACK.MAK) in the same subdirectory.  I wasn't referencing it, but when
I deleted it, things worked out fine.  As promised, and especially to
Dirk Lesko, Joe Booth, David Morgan and David Minter, here it is, a
great big:
        ĿĿĿĿ
          ͻ                              ͻ  
                 ͻ   ;ͻ   ;         
          Ľ                              Ľ  
          ͻ                              ͻ  
                                                    
                                                    
        ;     ;     ;          ;     ;     ;
              Ŀ ĿĿ    Ŀ Ŀ
                ͻ     ͻ             
                                   Ľ     
                Ľ           ͻ     
                ͻ                         
                                            ;
                Ľ     Ľ             Ŀ
              ;;    ;     ;


Msg#7954   To:ALL                                        
From:WALT BORYS                                 Date:02/03/89
Subject:NEW KID IN TOWN AGAIN                   Parent Msg#7954
------------------------------------------------------------------------------
garbled last message, lets try again- i'm new to this BBS, but pretty im-
pressed!  quick questions, though probably been answered before, but bear
with me-
1- Any utilities to pack or compress DBT files after DBF record deletes?
   could do it with F-func() or C udfs, but am afraid of network impli-
   cations.  is it worthwhile for me to reinvent more normal stuff, and
   upload the code?
2- When using the StoreToMemVars-EDIT-REPLtoFile strategy for editing
   records, what's the best strategy for networked access?  does the
   DBF have to be left open for RLOCK() to stay on? just entering the
   wunnerful world of networks...
3- Any body see the article on BulletProofing Network Code (DBMS or DBA)?
   Had the idea of 'Card-catalogue', where you 'sign-out' a record for
   write access into a 'catalogue' file of recno's (after first making
   sure its not already out.)  other users go thru similar logic.  any
   comment from the cognesciente out there on pros and cons?
  
Thanx,  WaltB


Msg#7956   To:CHRISTOPHER EDGAR                         (Rcvd)
From:JOE BOOTH                                  Date:02/03/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Give me a call sometime at work (215) 953-1202 and I'll describe a
technique which will work.  It's really quite simple, and work's fine.


Msg#7957   To:DIRK LESKO                                (Rcvd)
From:CHRISTOPHER EDGAR                          Date:02/03/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Dirk,
Do you mean usFUNCky' random() function in the index expression? How
would I use random() in this case?
 
Thanks
--Christopher


Msg#7958   To:WALT BORYS                                (Rcvd)
From:DIRK LESKO                                 Date:02/03/89
Subject:(R)NEW KID IN TOWN AGAIN                Parent Msg#7954
------------------------------------------------------------------------------
To Pack the DBT, just copy to <file> for ! deleted() and the DBT file
will be packed for you....The best bet for locking records would be to
either load the fields into memvars, and then release the lock, and edit
the memvars, or, use semaphore locks to control the locking, and lock
type (critical, volatile,inuse) so that your app can decide what action
to take, rather than have rlock() decide for you, you may be able to have
the record marked in use while someone is looking at it, then if someone
edits it, you can flag the edit so the person looking at it can be
updated onthe spot....etc..
 
dLESKO


Msg#7959   To:CONSTANTIN FLORESCU                       (Rcvd)
From:DIRK LESKO                                 Date:02/03/89
Subject:(R)SET KEYS                             Parent Msg#7959
------------------------------------------------------------------------------
Try a distributed network where each hard disk can be accessed by any
other. There is no server/station relationship. Something like TOPS is
rather inexpensive, and no server technology is necessary, but you can
share data/programs...
 
dLESKO


Msg#7960   To:CHRISTOPHER EDGAR                         (Rcvd)
From:DIRK LESKO                                 Date:02/03/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
you can use random() to choose a record number, just use the following
code to pick a record randomly:
 
        use <DBF>
        temp = random()
        do while temp > lastrec()
                temp = random()
        enddo
        goto temp
 
dLESKO


Msg#7961   To:ANDREW GROSS                               
From:JOHN THORN                                 Date:02/04/89
Subject:(R)WORKHORSE PRINTER                    Parent Msg#7644
------------------------------------------------------------------------------
I've used the HP DeskJet for about 9 months now.  The thing is great, so
long as the paper doesn't get wet (HP folks are working on non-water-
soluable ink).  It prints out at 75/150/300 dpi - no problems - can get
a full ream (full page text/graphics) out of a single $15 cartridge.


Msg#7962   To:CHRISTOPHER EDGAR                         (Rcvd)
From:JOHN THORN                                 Date:02/04/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Try Funcky!  Determine the seed....etc, etc!


Msg#7963   To:JOE BOOTH                                 (Rcvd)
From:RON LANDBERG                               Date:02/04/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
ALT-E is definitely 274.  I use it all the time for operator to invoke
edit mode.


Msg#7964   To:CONSTANTIN FLORESCU                       (Rcvd)
From:RON LANDBERG                               Date:02/04/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
You can definitely trap all thos ALT +alpha strokes.  I do it all the
time.  ALT-E is definitely 274.  Whatever is wrong, it's not a limit on
the key value.  I'll thread back and read your original message and see
if I can add anything more.


Msg#7965   To:CONSTANTIN FLORESCU                       (Rcvd)
From:RON LANDBERG                               Date:02/04/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Where are you trying to use this assignment?  If inside DBEDIT, don't. 
Instead of SETting the key, just check for LASTKEY()=274 inside your UDF
attached to your DBEDIT.  Then do whatever you would do in your doedit
procedure  --  right there inside the DBEDIT UDF (Or call DOEDIT from
THERE).
 
If you are not in DBEDIT, then you should propbably let us see the DOEDIT
proc to see if we can spot the problem.  I answered another message that
you could test for keys above 255, but maybe it's true that you cannot
SET a key jigher than 255.  I guess I'll try it when I get off.


Msg#7966   To:WALT BORYS                                (Rcvd)
From:RON LANDBERG                               Date:02/04/89
Subject:(R)NEW KID IN TOWN AGAIN                Parent Msg#7954
------------------------------------------------------------------------------
A lot really depends on the amount of access you expect to have for any
SPECIFIC record.  Anotherwards, if you have a database in which it is
VERY likely that two operators will often be trying to access and change
the same record at aout the same time, then you should strongly consider
a locking scheme which will lock the record as soon as the first operator
grabs it. Then keep it locked until the first operator releases all use
of the record.  The other operators still can read the record.  They just
won't be able to change it until the first operator is finished.
 
On the other hand, if it is unlikely that more than one operator will be
trying to access the same record at the same time, then just lock it when
the memvars are being written back to the record.  
 
Also realize that if a data change is keyed in and the changes will be
written out to 2 or 3 related files, by locking the record in the parent
file, and keeping it locked until all records in the related files are
updated, you never have to worry about locking the records in the child
files.  A second operator accessing the same record will not be able to
get past the lock on the record in the parent file.


Msg#7971   To:DIRK LESKO                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/04/89
Subject:(R)SET KEYS                             Parent Msg#7959
------------------------------------------------------------------------------
Thanks. I will do & let you know.
Speak to you.
<CF>


Msg#7972   To:RON LANDBERG                              (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/04/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Sure Ron... I use it also know.. after I found that instead of 
  set key 274 to expedit
I was saying
  set key -274 to expedit
Thanks
<CF>


Msg#7973   To:RON LANDBERG                              (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/04/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
Yehhh... I was doing something wrong there Ron.
Instead of 
set key 274 to expedit
I said
set key -274 to expedit
See it ? That "-" messed me up.
Thanks for your help anyhow.
<CF>


Msg#7975   To:NOEL MANZANILLA                           (Rcvd)
From:CLIFF GREEN                                Date:02/04/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Ah.  First off, what kind of print "tricks" are you looking to do, and do
you think you'll be trying for a HP LaserJet or another brand?  (I know,
I'm repeating myself, but there's reason behind it.)  If you're looking
at HP lasers, and you're going to be doing any programming for it, try to
convince your company to get the so-called Technical Manual for the durn
thing at the same time (I don't recall the part number offhand) -- it
contains more of the PCL code than the standard user's manual that comes
with the printer.
Generally, you'll find that these critters have a distinct language all
their own which you can easily use from within Clipper, and since they
are _page_ oriented printers, much more labile than your run-of-the-mill
line-oriented printer.
Again, as an example, if you want to design a form to be printed out on
each page (or not each) you can dump it to the printer's memory, then
tell the printer to either "switch it on" or "ignore it", so to speak. 
Forms, lines, moving around on the page in varying increments,
pushing/popping cursor locations....  The list goes on.  So, I return to
the original question -- What kinds of things do you anticipate needing
to do?  Incidentally, if you feel like calling some other bbs's, there
are a few with emphases on lasers and dtp;  Cooperworks (or whatever
they're calling themselves now) was at: 1-608-271-3685 the last time I
called (a few months ago), and there was another one here in NJ, the
FormFeed bbs, at: 1-201-869-1327 (again, it's been some time for me since
I last called there).  You can generally get some straight poop from the
folks there (especially Cooperworks), and then come back here for the
real skinny on making it work in Clipper.


Msg#7976   To:NOEL MANZANILLA                           (Rcvd)
From:CLIFF GREEN                                Date:02/04/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
I don't much like the machines with the Ricoh engine (too messy, and the
print doesn't fuse to the paper as well), and prefer the Canon engine
(like the HP), but that's just my opinion, and probably reflects the
lopsided experience I've had.  One of our people has an AST laser (I
believe), and swears by it -- it takes a lot of paper in its tray(s?),
can do raster and vector graphics, postscript, and HP emulation.  As I
mentioned, I've doodled with the Kyocera, but wasn't massively
impressed.  Nonetheless, if the machine you wind up with has the features
you want, you'll probably be very satisfied (how's that for tautology?).
Oh yeah, don't ask me about real-world cost comparisons:  I work for a
large institution and we get next-to-obscene discounts.


Msg#7978   To:DIRK LESKO                                (Rcvd)
From:FRANCOIS DESROSIERS                        Date:02/05/89
Subject:FUNCKY                                  Parent Msg#7978
------------------------------------------------------------------------------
Does Funcky has some functions that bypass the bugs with the memo fields
in Clipper ie strtran(), mlcount, etc. And does is has a function to
compress the DBt file that has been modified adn who has grown out of
limit.
 
                      Francois Desrosiers


Msg#7980   To:MARK NEIDORFF                             (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/05/89
Subject:(R)EMS FOR THE PC                       Parent Msg#7753
------------------------------------------------------------------------------
Oh no, you're thinking of Extended memory. The 808x can only access1MB of
memory, but the 286 on up can go beyond that 1MB limit. Memory beyond
640K and esp. 1MB is generally referred to as Extended.
 
On the other hand, Expanded memory is a whole other ballgame. Extended
memory can only be accessed from Protected mode, which is why 808x
machines don't have it: no protected mode. Expanded memory is done by
bank switching. It uses a 64K bank of memory (usually D000 - DFFF) to
swap data between the EMS card memory and the 'paged' RAM. EMS is
possible on a PC. Extended is not. The reason Excel PC requires a 286 or
better is due to several reasons: the raw computing power, and probably
when they wrote it, they decided it would need at least the power of a
286, so why not have it take advantage of the 286 instructions? Things
like Protected mode, 286 opcodes not available on the 8086, etc. Thus, no
PC for Excel PC.
 
AST and STB make EMS cards for the PC (and the AT of course.) There is
also a noname company that makes them, but beyond a price (believe it or
not, an EMS card for $95?!? Taiwan's a great country, no?) I didn't
pursue them. When the salesman said they don't have them in stock because
they were recalled due to a flaw in the design (aka major bug) I didn't
ask for more details...


Msg#7981   To:MARK NEIDORFF                             (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/05/89
Subject:(R)EMS FOR THE PC                       Parent Msg#7753
------------------------------------------------------------------------------
BTW, regarding EMS and Extended memory:
 
When the 286 chip was designed, the addressing would hit F000:... and on
a 8086 if you try to address something beyond that (how much beyond it, I
do not remember), it would wrap around to 0000:0000, but the 286 chip
would keep on addressing. IBM builds the AT, but they need to work around
this "bug" in the 286 design so the AT performs like the PC. If address
line 20 I believe it is (or pin 20, or something like that, but it's #20)
is high, the addressing acts like a PC, wrapping around if need be. If
it's low, addressing beyond 1MB is possible. (Note: don't quote me on the
exact setting, I don't remember which way is high and which is low).
However, the only time this line/pin/etc. is set to allow > 1MB
addressing is in Protected mode. Thus was born Extended memory. Note: The
286 chip was built by Intel, and IBM put it in the motherboard and wired
this line 20 so operations would mimic the PC. This is not Intel's fault,
since they just built the chip, which was supposed to access > 1MB RAM.
 
Here's the good part. You know how Windows gets an extra 64K on 286/386
machines? Microsoft figured out how to access this memory addressing
normally wraps past and they designed what they call XMS - eXtended
Memory Specification I believe is the full name. When you want to access
this "extra" 64K, the machine jumps into Protected mode, does whatever,
and jumps back. Very ingenious, although I wonder how the guy who figured
it out stumbled across it. BTW, it's not really 64K, it's slightly less,
but for more exact details you can pick up the XMS specs from most
bulletin boards. I think it floats under the name HIMEM.ARC, a text file
of the specs. 


Msg#7982   To:FRANCOIS DESROSIERS                        
From:DIRK LESKO                                 Date:02/05/89
Subject:(R)FUNCKY                               Parent Msg#7978
------------------------------------------------------------------------------
Not yet, although it has been a frequent request to do so. So maybe you
will here of something like that soon. As far as packinig a DBT, you can
just use a copy to <file> for not deleted() to pack the DBT, and Steve
Straley has a utility written in Clipper that will pack the DBT as well,
you might want to ask him...
 
dLESKO


Msg#7983   To:RICHARD HORWITZ                           (Rcvd)
From:MARK NEIDORFF                              Date:02/05/89
Subject:(R)BEGIN SEQUENCE                       Parent Msg#7926
------------------------------------------------------------------------------
Begin sequence / End construct:
Suppose that you had a set of code that you wanted to break out of
quickly and neatly.  Wouldn't it be nice to have a command that will get
you to the end of , and out of, the set of code in one statement without
anyother statements being evaluated?  That is a purpose of Begin Sequence
/Break / Endsequence.  Here is a quick example of what I mean.
 
 
*  code fragment to show begin sequence
* code not tested and probably has bugs
begin sequence
A_Number = 1
begin sequence
        do while .t.
           ? A_Number
           key=inkey(.1)
           if key = 27  && esc was pressed
              break
           endif
           A_number = A_Number + 1
       enddo
end
* more code to follow'
 
 
As soon as the user presses the Esc key, program execution will jump to
the line after the END statement.  This means that : the do while is
terminated, and that A_Number is not incremented after the Esc is sensed.
Does that help?
BeginSequence can also be nested.
     <<  Mark  >>


Msg#7986   To:MARK NEIDORFF                             (Rcvd)
From:MENACHEM BAZIAN                            Date:02/05/89
Subject:(R)USING EXPANDED MEM                   Parent Msg#7616
------------------------------------------------------------------------------
Where can I find it???? Is it on disk here???
 
Me3nachem


Msg#7987   To:STEVE STRALEY                             (Rcvd)
From:MENACHEM BAZIAN                            Date:02/05/89
Subject:(R)GENERAL                              Parent Msg#7651
------------------------------------------------------------------------------
Steve,
 
I downloaded it and will take a look at it.
 
I'll let you know what I think.
Menachem


Msg#7988   To:RON HOCKEMEIER                            (Rcvd)
From:MENACHEM BAZIAN                            Date:02/05/89
Subject:(R)DISK FORMATTING                      Parent Msg#7708
------------------------------------------------------------------------------
Thanks ron,
 
I'll take a look at it.
Menachem


Msg#7990   To:MARK NEIDORFF                             (Rcvd)
From:RICHARD HORWITZ                            Date:02/05/89
Subject:BEGIN SEQUENCE                          Parent Msg#7990
------------------------------------------------------------------------------
 So it's just a quick way to jump to the end of a piece of code. Thanx.


Msg#7991   To:DIRK LESKO                                (Rcvd)
From:STEVE BADARACCO                            Date:02/06/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Dirk,
My problem occurs when I DON'T use OVERLAY PROG,$CONSTANTS!
If I omit the line entirely, or if I use OVERLAY CODE,$CONSTANTS, either
way will hang my application.  I have NO Autumn '86 code in any of my
OBJs or LIBs.
 
Granted, the use of OVERLAY CODE,$CONSTANTS produces a smaller load
module (by about 30K), but it hangs, while the other runs perfectly.
Remember the error I was getting a month ago ... the message "graoup
DGROUP is too big"?  That was on another S'87 application with overlays,
and the message persisted until I used "OVERLAY PROG,$CONSTANTS".
This is nuts; I think there is some serious bug un the PLINKer.
 
...Steve


Msg#7993   To:DIRK LESKO                                (Rcvd)
From:STEVE BADARACCO                            Date:02/06/89
Subject:PACKMEMO                 
------------------------------------------------------------------------------
Sorry, Dirk, I didn't intend that message to be private, so here it is
for everyone:
 
I'm uploading packmemo.arc, a very clean memo packer which I found on a
BBS (if it isn't here already).  Uses file I/O from EXTEND.LIB, and it
seems to be very clean.
 
Try it, you'll like it!
PS - It's callable from your programs because it's a PRG.
...Steve


Msg#7994   To:TERRY DILLARD                             (Rcvd)
From:DAVID MINTER                               Date:02/06/89
Subject:(R)ATTABOY!                             Parent Msg#7952
------------------------------------------------------------------------------
Shucks, thanx!


Msg#7995   To:DAVID MINTER                              (Rcvd)
From:STEVE BADARACCO                            Date:02/06/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
David,
Here is a detailed description of what happens:
  OVERLAY PROG,$CONSTANTS         No problem, 289K load size.
   
  OVERLAY CODE or                    Load size 267K, but ...
  OVERLAY CODE,$CONSTANTS         Freezes as the application is loading.
  
  <NO "OVERLAY" statement>        289K load size, loads Ok and runs for
                                   a while, then develops wierd memory 
                                   errors, such as FORGETTING VARIABLES
                                   and drawing random junk on the screen,
                                   then hangs.
...Steve


Msg#7996   To:ALL                                        
From:STEVE BADARACCO                            Date:02/06/89
Subject:PACKMEMO                 
------------------------------------------------------------------------------
Yes, PACKMEMO.ARC is already on this board.
I have had great success with it.  It returns 12 different error codes to
indicate the success of the operation (or failure, of course!).
 
PACKMEMO not only helps you understand how DBT files are built, but is
also a very good example of how to use the low level file I/O routines in
EXTEND.LIB.
 
I'm not getting paid for this ad, in case you wondered.  PACKMEMO is
PUBLIC DOMAIN, not shareware.
 
...Steve


Msg#7997   To:STEVE BADARACCO                           (Rcvd)
From:DIRK LESKO                                 Date:02/06/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Have you tried a call to Nantucket on this? (not that I think that they
will know what the problem is...) I have seen NUMEROUS complaints about
linking with overlays, and not all are related to add in libraries,
although some are (IDL for one...) you really have to work to make the
overlay scheme work....You need to know not only what your own code
calls, you also need to know what the UDFs you use call as well. All you
need is one piece of code to end up in a different overlay and you can
run into problems...unfortunately nowhere is there to be found the
DEFINATIVE GUIDE TO OVERLAYS (Hint Steve)....I myself avoid them like the
plague...
 
dLESKO


Msg#7999   To:STEVE BADARACCO                           (Rcvd)
From:ROB HANNAH                                 Date:02/06/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Am I missing something or isn't the S87 overlay statement supposed to be
OVERLAY CODE (but not $CONSTANTS)?
 
ROB


Msg#8001   To:DIRK LESKO                                (Rcvd)
From:ROB HANNAH                                 Date:02/06/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Assuming you can't pass a max value to random(), wouldn't the following
be quicker / more elegant:
 
    use <DBF>
    goto random() % lastrec() + 1
 
ROB


Msg#8002   To:RON LANDBERG                              (Rcvd)
From:JOE BOOTH                                  Date:02/06/89
Subject:(R)SET KEYS                             Parent Msg#7845
------------------------------------------------------------------------------
I know it is, I was just wondering if something was funny with the BIOS
or the keyboard.....


Msg#8003   To:ROB HANNAH                                (Rcvd)
From:DIRK LESKO                                 Date:02/06/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Maybe, but the mod of the lastrec() is always going to be a smaller
number than usual..
 
dLESKO


Msg#8004   To:CLIFF GREEN                               (Rcvd)
From:NOEL MANZANILLA                            Date:02/06/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
thanks for the info Cliff, will try out these boards.  I have been trying
to automate a few office forms, one of which has 80 lines on each page
and has the most number of irregular line spacings I have ever seen. It 
also has several columns and little boxes that are not spaced evenly of
course ! Working with a dot matrix printer certainly doesn't make my life
easier.  I would like to (eventually) generate the form as it is being
filled out. Possible ? thanks again for your input.
Noel


Msg#8005   To:CLIFF GREEN                               (Rcvd)
From:NOEL MANZANILLA                            Date:02/06/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Cliff, I didn't know AST had laser printers.  Thats something to look
into.  Thanks again.
noel


Msg#8006   To:STEVE BADARACCO                           (Rcvd)
From:ROBERT HEYMAN                              Date:02/06/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Steve.   I've been using the construct "Overlay Code" since the inception
of S87 without a problem......   just thought you should know
Rob Heyman


Msg#8008   To:MENACHEM BAZIAN                           (Rcvd)
From:MARK NEIDORFF                              Date:02/06/89
Subject:(R)USING EXPANDED MEM                   Parent Msg#7616
------------------------------------------------------------------------------
Sorry Menachem, headroom is a commercial product.  Just reviewed in PC
Magazine this issue.
     <<  Mark  >>


Msg#8009   To:RICHARD HORWITZ                           (Rcvd)
From:MARK NEIDORFF                              Date:02/06/89
Subject:(R)BEGIN SEQUENCE                       Parent Msg#7990
------------------------------------------------------------------------------
Yes Begin sequence/break/end will just jump to the endo f a piece of code
that is in a single loop.  But if you have multiple loops, and multiple
begin sequences, you have more control and power.  Still no great deal,
but it is nice to be able to write it in 3 lines of code instead of what
it would take normally.
     <<  Mark  >>


Msg#8010   To:DIRK LESKO                                (Rcvd)
From:ROB HANNAH                                 Date:02/07/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
If lastrec() is 2000, any number % lastrec() will be between 0 and 1999. 
Adding 1 gives a number in the correct range.
 
ROB


Msg#8012   To:ROB HANNAH                                (Rcvd)
From:DIRK LESKO                                 Date:02/07/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
Correct...
 
dLESKo


Msg#8015   To:DIRK LESKO                                (Rcvd)
From:ROB HANNAH                                 Date:02/07/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
On further thought, what is the max value that random() can give.  I
think that what you mean by skewing the results to the lower numbers
somewhat is that if you have a database with 55k records and random()
gives 0 - 64k-1 that that will happen.
 
Sorry about the misunderstanding...
ROB


Msg#8017   To:ROB HANNAH                                (Rcvd)
From:DIRK LESKO                                 Date:02/07/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
that's right, If I remember correctly, the max value random() will return
is 65535.
 
dLESKo


Msg#8018   To:CHRISTOPHER EDGAR                         (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/07/89
Subject:(R)RANDOM RECORDS                       Parent Msg#7944
------------------------------------------------------------------------------
I told you....
In fact I personally think is THE GREATEST !
<CF>


Msg#8019   To:ALL                                        
From:GEFF PURCELL                               Date:02/07/89
Subject:RECORD CORRUPTION PROBLEM               Parent Msg#8019
------------------------------------------------------------------------------
        We are experiencing a very frustrating problem with record 
corruption in a database application.  The systems are being run on IBM
PS/2 Model 50z's. The problem is this:
        Randomly, fields are being replaced in an existing record.
        This problem may effect just one record or at times several 
        records in a row (in record number order not index order).
        The problem does not show up until a rebuild of the index is
done.
        When we rebuild we use INDEX on <key> TO <filename> we do NOT use
        the reindex statemen.
        This problem did not show up for several months
        We have set the E in the SET CLIPPER statment to E000 to make
sure
        no expanded memory is being used.
        Whem a field gets replaced is contains information from another
        record.  The information is good info but it is in the wrong
record
        It has happened to records even though apparently the user had
not
        edited either record recently.
        The application is fairly large. It did have overlays, but we
        have elimanted them in hopes that that would clear up the problem
        but it has not.
        The database in question has approx 500 to 1000 records depending
        on the user.
        The data base also has 6 indexes.  It used to have 7.
        We are using IBM DOS 3.3
        There are no TSRs in memory and the config.sys contains the
following:
        FILES = 31
        BUFFERS = 8
        DEVICE = ANSI.SYS
        DEVICE = DASDDRVR.SYS
        We tried eleminating the DASDDRVR.SYS but this did not seem to
help so we put it back in as this is a fix to DOS 3.3 for the FORMAT
command (will allow you to format multiple disk in a row with out bombing
out)
                                Any help would be grealy appreciated
                                as this problem has been a pain in the
                                A** for about 3 months now.  And all of
                                the users are women who can COMPLAIN
                                LOUDDDDLLYYY!!!
                Geff Purcell and Bill Roundcount
                        Sigma Chemical - St Louis


Msg#8020   To:DAVE LOBEL                                 
From:CONSTANTIN FLORESCU                        Date:02/07/89
Subject:HELLO                    
------------------------------------------------------------------------------
Hello Dave ! See you on the other nod. How are you. Havn't see you on the
board for quite a while. You've been away ?
<CF>


Msg#8022   To:STEVE BADARACCO                           (Rcvd)
From:STEVE DAVIES                               Date:02/07/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
If you haven't already done so, you might want to get a copy of the new
snap and run it. The headers will list calling and called routines. This
proved very helpful in nailing down just what is calling what. It will
also echo the headings to seperate file if you want. The cross reference
table is just as helpful in this matter. Next to each function or proc
name it list what file it's from and below that it lists all references
to that name, giving the file and line number.
Steve


Msg#8023   To:ALL                                        
From:HOWARD KAPUSTEIN                           Date:02/07/89
Subject:FREE DISK SPACE 1.5      
------------------------------------------------------------------------------
Come and get it while it's hot. Free Disk Space 1.5 is hot off the
presses. New features: Requires only 15K of memory to run, network
compatible, DOS environment variable for easily modified default options
(if you don't want to recompile. Yep, full Turbo C 2.0 source is
included). Look for FREE15.ARC in a download section near you.


Msg#8025   To:GEFF PURCELL                              (Rcvd)
From:RICHARD HORWITZ                            Date:02/07/89
Subject:(R)RECORD CORRUPTION PROB
------------------------------------------------------------------------------
 I had a problem with the 50 a few months ago using dbedit. SOMETIMES it
would bomb on return from the UDF, and others not. I had the exact same
config.sys and autoexec.bat on it that I had onmy 80, and the 80 NEVER
had a problem (with CLIPPER anyway!). Same program and everything, never
did find the cause.....maybe IBM did something screwy with the 50?


Msg#8026   To:GEFF PURCELL                              (Rcvd)
From:ROB HANNAH                                 Date:02/08/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
What is the index key you use that blows the app?  Does it happen on any
index statement?  And have you tried a stand-alone 3 or 4 line index
program to see if it works?
 
ROB


Msg#8027   To:ROB HANNAH                                (Rcvd)
From:GEFF PURCELL                               Date:02/08/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
        The index key was something like this: UPPER(VAR) + UPPER(VAR2) +
UPPER(VAR3)  (I know this could be simplified) but anyway we may have
found the prolem this morning. When I checked his Libs he was using the
orginal S87 (with no time) Libs so we put the new Libs in his dir and
have recompiled the application.  The problem was so random that we could
not get it to appear here on a 50z or on our 80s. So we are just goint to
send out the app to the users and see what happens now with the new lib
version.
Thanks for the suggestions,                     Geff Purcell


Msg#8028   To:GEFF PURCELL                              (Rcvd)
From:JOE BOOTH                                  Date:02/08/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
Is this a multi-user system, or single user ?


Msg#8029   To:STEVE BADARACCO                           (Rcvd)
From:DAVID MINTER                               Date:02/08/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Steve,
 From what I've allways been using, OVERLAY CODE, it has always worked
for me.  The only thing I can think of is that you are having a thrashing
problem somewhere.   I noticed that you use funcky.  One of the problems
that I found (Sorry Dirk) was trying to use ondo() in a lower level
overlay to call something in a higher level overlay, and ondo() was
declared EXTERNAL in the top level.  Took me 2 days to figure it out. 
Could something like this be happening to you? 
 
Dave


Msg#8030   To:CONSTANTIN FLORESCU                       (Rcvd)
From:ANDREW GROSS                               Date:02/08/89
Subject:WORKHORSE PRINTER                       Parent Msg#8030
------------------------------------------------------------------------------
Not attempting to beat a workhorse printer to death...
 
The dmp I use is a Toshiba P321SL.  It's fame-to-claim is that all
parameter changes (emulation, speed, etc.) is done via front panel
operations using a touch pad and an LCD window...it's not as advanced as
a DeskJet, but it's an easy to live with dmp...and unlike its
predecessor, it has an IBM proprinter emulation built-in.
 
andrew g.


Msg#8031   To:NOEL MANZANILLA                           (Rcvd)
From:ANDREW GROSS                               Date:02/08/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
If you do find a cheaper alternative, be reassured from an independent
source that the printer is reliable.  I've dealt a small bit with HP
"compatibible" printers...I prefer the real-thing.  Besides, who has yet
cloned the IID <grin> ?
 
andrew g.


Msg#8032   To:CLIFF GREEN                               (Rcvd)
From:ANDREW GROSS                               Date:02/08/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
The HP Technical Manual for the LaserJet II is Manual Part No.
33440-90905.
 
One other advantage to the II, my company recently bought a Label CAD
system for the LaserJet II.  The hardware part of it looks like a font
cartridge, tastes like it, and smells like it.  But it ain't!  When the
cartridge is plugged-in and the printer turned on, it takes over control
of the 68000 and from there on out all the messages that appear in the
LCD window and the performance of the indicator lights are specific to
the CAD package.  (The CAD package is used for barcode labels, but not
just the ordinary ones.  The packages makes EAN, A/2, E/2, CodaBar, etc.)
 
andrew g.


Msg#8033   To:NOEL MANZANILLA                           (Rcvd)
From:ANDREW GROSS                               Date:02/08/89
Subject:MSG 8032                 
------------------------------------------------------------------------------
Please seem message 8032.  I meant to route it to you.
 
andrew g.


Msg#8034   To:GEFF PURCELL                              (Rcvd)
From:DIRK LESKO                                 Date:02/08/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
Wouldn't upper(var1+var2+var3) be a little faster?


Msg#8035   To:DAVID MINTER                              (Rcvd)
From:DIRK LESKO                                 Date:02/08/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Hi dave, don;t worry about being sorry, ondo() calls the function via
Clipper's macro parser so the linker never knows what function it is
supposed to link and where or why...An explicit EXTERNAL call must be
declared to make sure that it gets put into the right module....Same goes
for execute() and timeout().....Also, the high level mouse fucntions call
lots of lower level functions that must be taken into account....
 
dLESKO


Msg#8036   To:ALL                                        
From:STEVE DAVIES                               Date:02/08/89
Subject:NEW (OLD) FILES                         Parent Msg#8036
------------------------------------------------------------------------------
Isn't it just *&/*%!@ wonderful when you download some nice new file,
unarc it and find it's some old thing re-arced because someone felt
obligated to upload something. The file I refer to is called CLIP_BRO.ARC
and contains a routine dated july 1986. Waste my time why don't you!


Msg#8037   To:NOEL MANZANILLA                           (Rcvd)
From:CLIFF GREEN                                Date:02/08/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Of the irregular line spacings and columns and boxes which are not spaced
evenly, are there any which depend for their spacing on the data to be
printed?  Put a bit differently, are these irregularly spaced items
_always_ at the same place on the sheet?  If so, then a template would
seem to be a good answer to your dilemma, and a laser will do that (at
least HP's will, and I assume most if not all others will -- they'd be
dumb to leave something like that out).  You can even pretty up your
forms with variable-width lines, a variety of gray shading, and as many
dingbats and bells and whistles as your fontware (how's that for a
neologism?) will allow.  What you would wind up doing in that case, is:
1) design your form, and code the template.  It gets downloaded to the
printer once per session (don't turn the power off!), and turned on or
off as you need it -- this takes up printer memory, so get enough!
2) using your design (see #1), write the code to take your data and stuff
it into the empty spaces on the page.  Just because the form goes to 80
lines doesn't mean the page will be ejected (at least, not if you've set
the logical length of the page to allow for this - you'll need to delve
into VMI - vertical motion indices, and suchnot).  It sounds like you've
got some interesting exploration to look forward to.


Msg#8038   To:ANDREW GROSS                              (Rcvd)
From:CLIFF GREEN                                Date:02/08/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
(re:  Label CAD cart)   Nifty!
 
Do you know if there are any other products that take over the LJII like
that?  There's got to be some potential there for a variety of
applications (e.g., vector graphics in firmware, rather than software, so
it won't take up valuable memory, etc.)


Msg#8039   To:STEVE DAVIES                              (Rcvd)
From:MARK NEIDORFF                              Date:02/08/89
Subject:(R)NEW (OLD) FILES                      Parent Msg#8036
------------------------------------------------------------------------------
Yeah, isn't that the file that is good for A86 only?
     <<  Mark  >>


Msg#8040   To:CURTIS LITTLE                             (Rcvd)
From:JOHN NEWTON                                Date:02/08/89
Subject:SWITCH PROBLEM                          Parent Msg#8040
------------------------------------------------------------------------------
Curtis,
 
I've been trying to get SWITCH to work with my application. 
Unfortunately
I haven't had much luck. Every time I get to the SWITCH! function I get
the
SWITCH! banner across the top of the screen then it locks up with a
blinking
cursor in the upper left corner.
 
The following is my make file:
 
         Main.obj  : main.prg  
           clipper main -m
 
         genproc.obj : genproc.prg
           clipper genproc -m
 
         CONOPS.EXE : main.obj genproc.obj
         PLINK86 @M
 
The following is the M.LNK file
 
         OUTPUT conops
         FI MAIN                     - the main shell
         FI E:\OBJ\SWITCH87          - Switch! module
         FI GENPROC                  - All functions & procedures
         FI E:\OBJ\FUNCKY            - Dirk's FUNCky
         FI E:\OBJ\EXXTEND           - used by DMS' error routines
         FI E:\OBJ\PRT_SCRN          - Print screen routine
         LIB E:\LIB\CLIPPER         
         LIB E:\LIB\FUNCKY
         LIB E:\LIB\EXTEND
 
I've also tried my luck with Borland's TLINK, but again no luck.
 
        tlink @flist,conops,, $(lib1) $(lib2) $(lib3)
 
flist:
        MAIN+
        e:\obj\switch87+
        genproc+
        e:\obj\funcky+
        e:\obj\exxtend+
        e:\obj\prt_scrn
 
The procedure I'm using is as follows:
 
  Procedure UPDT
1 if !cur_month
2    BEGIN SEQUENCE
3      * Check and see if PC-DATA exist -
4      * If so rename it to PC-DATA.BAK
5      ********************************
6      if isfile("PC-DATA.TXT")
7        rename PC-DATA to PC-DATA.BAK
8      endif
9       
10     * Switch out of program to ZCOMM and download data
11     **************************************************
12     switch87("ZCOMM")
13     if !isfile("PC-DATA")
14       print(23,1,"ERROR! Data file not created! Press return to
continue")
15       set console off
16       wait
17       set console on
18       BREAK  
19     else
20       rename PC-DATA to PC-DATA.TXT
21     endif
22
23     t_dbf = substr(cmonth(date()),1,3) +
substr(ltrim(str(year(date()))),-2)
24     t_ntx = t_dbf
25     t_loc = substr(t_dbf,1,3) + "LOC" + substr(t_dbf,4,2)
26     t_prj = substr(t_dbf,1,3) + "PRJ" + substr(t_dbf,4,2)
27     use &M_dbf
28     copy structure to &t_dbf
29     use &t_dbf
30     append from PC-DATA.TXT SDF
31     do sel_dbf
32     use &M_dbf
33     do ntx_test
34          do array_set
35        END
36 endif
37 return
 
The procedure makes it to line 12 and I get the SWITCH! message then it
locks
up with a blinking underline cursor in the upper lefthand corner of the
screen.
The demo programs work fine, but mine just hang. Sure hope you can point
me in
the right direction 'cause I'm lost.  
 
Thanks, John Newton



Msg#8042   To:ANDREW GROSS                              (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/09/89
Subject:(R)WORKHORSE PRINTER                    Parent Msg#8030
------------------------------------------------------------------------------
How fast and how many copies ?
Thanks.
<CF>


Msg#8045   To:STEVE DAVIES                              (Rcvd)
From:DIRK LESKO                                 Date:02/09/89
Subject:(R)NEW (OLD) FILES                      Parent Msg#8036
------------------------------------------------------------------------------
Yeah, that's a problem...I have done that several times myself, and on
some BIG files....
 
dLESKO


Msg#8046   To:JOHN NEWTON                               (Rcvd)
From:DIRK LESKO                                 Date:02/09/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
Ask him what public symbols switch may be trying to read...maybe FUNCky
has one that satisfies it and causes a conflict???? 
 
dLESKO


Msg#8048   To:DIRK LESKO                                (Rcvd)
From:JOHN NEWTON                                Date:02/09/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
As I gather from previous messages, there didn't seem to be any, but
leave it to me to find such.  
  
I did notice in the doc file that when he complied, he had the programs
backwards, or so it seemed to me. I probably need some lessons in the
fine art of compiling anyway.
 
-John


Msg#8049   To:GEFF PURCELL                              (Rcvd)
From:RON HOCKEMEIER                             Date:02/09/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
I've had corruption problems due to power supplies and static.
The problem closest to your's occured when the computer was inadvertently)
(inadvertently) on the same power circuit as a photocopy machine: random
field corruption (which coincided to writing to disc at the same moment a
photocopier was used).  I put a VOM on the circuit and found the voltage
dropping as low as 80VAC.
We've also lost files--and had corruption--due to static. (You could stand
6" from the computer and feel the "static arc."  We couldn't fix it;
fortunately we moved.
Unfortunately, the new location had at least one bad power outlet. 
We lost about 10% of the FATs on floppies...
We solved that by installing an on-line, uninterruptible power supply.
Hope this helps.


Msg#8051   To:ALL                                        
From:HARRY GILBERT                              Date:02/09/89
Subject:ARRAYS                                  Parent Msg#8051
------------------------------------------------------------------------------
Help! I'm having problems getting ACHOICE() to work with arrays
containing numeric data. An example follows:
   DECLARE names[FCOUNT()],type[FCOUNT()],size[FCOUN()]
   AFIELDS(names,type,size)
   pick = ACHOICE(10,10,20,20,names,.T.)
   pick = ACHOICE(10,10,20,20,type,.T.)
   pick = ACHOICE(10,10,20,20,size,.T.)
.
I can view the character array elements in "names" and "type", but the
pick-list for "size" is blank. But if I do:
   FOR x = 1 TO FCOUNT()
     ? size[x]
   NEXT
the field lengths are there!
Also, if I AFILL(array,0), an array should be filled with the number 0,
but again, an ACHOICE on the array comes up blank. What'm I doing wrong.
If ACHOICE won't work on numeric data, I'm in real trouble. I've tried
both TLINK and PLINK86 thinking it's related to "floating point" numbers,
to no avail.


Msg#8052   To:DIRK LESKO                                (Rcvd)
From:GEFF PURCELL                               Date:02/09/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
Yes it would, but I did not build this system and the Guy who did must
not realize that.  I told him once or twice. (oh well some people are
just slow to learn).
        Geff


Msg#8053   To:RON HOCKEMEIER                            (Rcvd)
From:GEFF PURCELL                               Date:02/09/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
We think we found the problem now... the guy who wrote the system had the
old Sum 87 libs so we gave him copies of the new libs and everything
seems to be working know.  You idea would have been good to check but the
22 users are all in dif states. 
Thanks for the idea though                      - Geff


Msg#8054   To:ANDREW GROSS                              (Rcvd)
From:NOEL MANZANILLA                            Date:02/09/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
thanks, saw the message
noel


Msg#8055   To:ALAN TOMPKINS                             (Rcvd)
From:NEIL WEICHER                               Date:02/09/89
Subject:(R)CLIPPER RUNTIME ERROR 
------------------------------------------------------------------------------
Are you trying to reindex from two stations at once?  
 
Regards,
Neil


Msg#8056   To:MARK NEIDORFF                             (Rcvd)
From:MENACHEM BAZIAN                            Date:02/09/89
Subject:(R)USING EXPANDED MEM                   Parent Msg#7616
------------------------------------------------------------------------------
Okie dokie. I will look at it and lket you know.
 
Thanks a mil.
Menachem


Msg#8057   To:HARRY GILBERT                             (Rcvd)
From:DIRK LESKO                                 Date:02/09/89
Subject:(R)ARRAYS                               Parent Msg#8051
------------------------------------------------------------------------------
Harry, achoice() only works on strings....sorry to give you the bad
news....Why not create an array that fills up with str() of the numbers?
 
dLESKO


Msg#8058   To:HARRY GILBERT                             (Rcvd)
From:MARK NEIDORFF                              Date:02/09/89
Subject:(R)ARRAYS                               Parent Msg#8051
------------------------------------------------------------------------------
Can't convert the numbers to strings just for the achoice?
     <<  Mark  >>


Msg#8059   To:STEVE DAVIES                              (Rcvd)
From:ROBERT HEYMAN                              Date:02/09/89
Subject:(R)NEW (OLD) FILES                      Parent Msg#8036
------------------------------------------------------------------------------
here here !


Msg#8060   To:CLIFF GREEN                               (Rcvd)
From:ROBERT HEYMAN                              Date:02/09/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Cliff....HP templates.....  I've known about the existance of such a
thing...but have never been able to find out precisely HOW to download a
template (such as a form) to the laser.  Can you be more specific as to
how that is done... and how particularly it might be accomplished from
within clipper.
Thanks
Rob Heyma


Msg#8061   To:DIRK LESKO                                (Rcvd)
From:HARRY GILBERT                              Date:02/09/89
Subject:(R)ARRAYS                               Parent Msg#8051
------------------------------------------------------------------------------
I was hoping to use the arrays in a package that analyzed surveys, and
processed the database, summing the responses for a report. Oh, well. How
come there's no mention of achoice being limited to strings in Clipper's
docs? Maybe they should get you to write their next docs. At least we'd
KNOW what's going on. Thanks.


Msg#8062   To:MARK NEIDORFF                             (Rcvd)
From:HARRY GILBERT                              Date:02/09/89
Subject:(R)ARRAYS                               Parent Msg#8051
------------------------------------------------------------------------------
I guess I'll have to, but the goal was to peek into some survey data,
weed out "flyers" and go on with processing. Having parallel arrays of
numeric and string data would eat up too much RAM.


Msg#8064   To:RON LANDBERG                              (Rcvd)
From:STEVE DAHLIN                               Date:02/10/89
Subject:(R)NOVELL 2.0A LIMIT                    Parent Msg#7759
------------------------------------------------------------------------------
Thanks all for the help.  Using the patch I can now access more than the
40 file limit.


Msg#8067   To:DIRK LESKO                                (Rcvd)
From:STEVE BADARACCO                            Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Guess I'll try Nantucket.
But if nobody on this board has the answer, I haven't much hope for
getting one from them!!!!!
..Steve


Msg#8068   To:ROB HANNAH                                (Rcvd)
From:STEVE BADARACCO                            Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
You're right, Rob.
The point is, none of the "formats" work, except OVERLAY PROG,$CONSTANTS.
That's the wierd part.  I cannot leave the line out entirely, either, or
the application hangs.
...Steve


Msg#8069   To:DAVID MINTER                              (Rcvd)
From:STEVE BADARACCO                            Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Thanks, Dave.
However, I do not use ondo() at all in the application.
...Steve


Msg#8072   To:STEVE BADARACCO                           (Rcvd)
From:DIRK LESKO                                 Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
I guess that's a compliment for the board....
 
dLESKO


Msg#8073   To:ALL                                        
From:DAVID MINTER                               Date:02/10/89
Subject:CLIPPER 89 DEV CONFERENCE               Parent Msg#8073
------------------------------------------------------------------------------
 
It's official!  1989 Nantucket Clipper Developers Conference
 
       Sponsor: Nantucket Corporation
          When: June 11-14, 1989
         Where: Los Angeles, California, at the Airport Marriot
      How many: Attendance limited to 750
Registration &
   information 
       inquiry: 1-800-848-CLIP (1-800-848-2547) Monday - Friday
          Cost: $595 per person before May 11, 1989
                $695 per person after May 11, 1989
 
The 1989 Clipper Developers Conference will be an opportunity for
Clipper Developers, programmers and users from around the world to
gather together to exchange ideas, learn about emerging directions
for Clipper, participate in discussions about future directions, gain
advanced insight into Clipper, attend numerous technical information
sessions, visit the only Clipper-specific trade show in the world,
and enjoy fine food, drink, and conversation.  The conference
provides a unique opportunity to meet key Nantucket employees, third
party add-on developers, and other interested Clipper users from
diverse backgrounds.  
 
Last year's Clipper Developer Conference was a tremendous success
-- this year's will be even better.  Call today to receive more
information and registration materials.  Last year's conference
sold out.
 
Stay tuned for future bulletin updates (scheduled speakers, seminars,
etc).


Msg#8076   To:JOHN NEWTON                               (Rcvd)
From:CURTIS LITTLE                              Date:02/10/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
John,
  I was standing by Ken when he answered, but just wanted to make sure
you knew I was answering.  The older one had a problem in the interface
between Clipper and the actual Switch! assembly function it was weird, it
didn't show up in the applications I use Switch! in, but someone sent me
some code that demonstated the problem and was what enabled me to fix it.
The fixed one is in SWITCH22.ARC in dir 99.


Msg#8078   To:DIRK LESKO                                (Rcvd)
From:STEVE BADARACCO                            Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Yes, it was intended to be.
...Steve


Msg#8079   To:DIRK LESKO                                (Rcvd)
From:JON MELTZER                                Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Isn't someone from Nantucket on this board?


Msg#8080   To:ALL                                        
From:JON MELTZER                                Date:02/10/89
Subject:RTLINK                                  Parent Msg#8080
------------------------------------------------------------------------------
If anyone here has used the RTLink linker from Pocket Soft (advertised
on page 85 of the February 1989 "Data Based Advisor"), I would like to
hear your evaluation of the product. Thank you.


Msg#8081   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/10/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Bob,   let me work up a simple sample over the weekend (gawd! my wife's
gonna love that!), and I'll ul it later.


Msg#8082   To:HARRY GILBERT                              
From:MARK NEIDORFF                              Date:02/10/89
Subject:(R)ARRAYS                               Parent Msg#8051
------------------------------------------------------------------------------
Yes, I see that parallel arrays would eat ram, but couldn't you convert
them to string on the way into the array?
     <<  Mark  >>


Msg#8083   To:JON MELTZER                               (Rcvd)
From:DIRK LESKO                                 Date:02/10/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Supposedly.....They check in from time to time..
 
dLESKO


Msg#8084   To:JOHN NEWTON                               (Rcvd)
From:CURTIS LITTLE                              Date:02/11/89
Subject:SWITCH!                                 Parent Msg#8084
------------------------------------------------------------------------------
Just wanted to let you know that I'm sending a new one tonight.  If you
already grabbed version 2.2 somewhere then the problem you were having
would be fixed.  If you have 2.2 there's no need to get 2.3 because the
main changes I made in 2.3 are in the documentation.  The docs for 2.3
fixed the backwards param passing you mentioned to Dirk (in the ASM
example where the offset was pushed before the segment).  In addition the
new docs show the proper way to interface Switch! from BASIC.  The only
functional change to the function from 2.2 to 2.3 is that beginning 2.3
the temp file is no longer created as a hidden system file, instead it is
simply a read-only file.  Thanks for mentioning the part about the
example being in error - that led to the corrected example in the
appendix.


Msg#8085   To:DIRK LESKO                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/11/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Dirk, don't ask me about my experience with Nantucket Tech Support or...
let's not speak about..AT. Most of their techs don't know too much...and
besides you have to wait a lot on hold to find out that in 99% of cases
you know more than they know(of course I don't include here the heavy
weight who actually wrote CLipper...But who can ever speak to them except
probably...You)
<CF>


Msg#8086   To:KEN CAMP                                  (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/11/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
Ken, let me jump in for myself and ask you a question. Does Curtis work
for you or you work for him. Just wanted to know...
<CF>


Msg#8087   To:JON MELTZER                               (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/11/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Jon, about a month ago a lady came on this board(I would say pretty
strong) and told everyone that she will log in every week...I didn't see
any messages(at least public) since then...
<CF>


Msg#8088   To:CONSTANTIN FLORESCU                       (Rcvd)
From:DIRK LESKO                                 Date:02/11/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
The board was probably more active than she could handle.....Dave Morgan
checks in on a regular basis, but it is usually every 2-3 weeks so he
must have a handful when he logs on...
 
dLESKO


Msg#8089   To:ALL                                        
From:DIRK LESKO                                 Date:02/11/89
Subject:DBMS OR NOT....                         Parent Msg#8089
------------------------------------------------------------------------------
As moderator, I do like to bring to everyones attention items of
interest. This text is contained in the file pc_week.arc and it
asks us to respond. I am willing to pass along this thread to the
persons mentioned at the end of this letter if you are all willing to
participate openly in this forum. I believe that this is a rather
interesting topic that should generate a lot of response and debate.
So, Read the letter, and the challenge, and at the end of the message I
will propose a few questions that you may respond too.
-----------------------------------------------------------------------
        
Dear Clipper User:
         
Clipper was not included in the recent PC Week Poll of Corporate 
Satisfaction for programmable and relational DBMS systems (Volume 
6, Number 2  January 16, 1989) or in their Buyer's Guide for 
Database Management Systems (Volume 6, Number 4 January 30, 
1989).  When Nantucket's public relations agency contacted John 
Helliwell, Executive Editor at PC Week and asked why Clipper was 
not included,  he replied with the attached letter which says Clipper 
is "a wonderful AUXILLIARY product to a database management 
system, but is not itself a DBMS."  He is correct in supposing that we 
are going to dispute his conclusion.
        
We would like you to participate in the test he proposes.   He asks to 
be put in touch with Clipper users who DO NOT use it in conjunction 
with dBASE or another full-fledged DBMS.
       
If you are using Clipper as a programmable DBMS to develop 
relational database applications, please send a description of how you 
use Clipper and the types of applications you create to:
      
John Helliwell
Executive Editor
PC Week 
800 Boylston St.
Boston, MA  02199
        
Please include your phone number if you are willing to be contacted 
by the magazine.   He might also be interested in the Clipper add-on 
products you use as part of your development environment.
If you respond by mail, MCI or FAX, please send a copy of  your 
response to:
         
Judy Mason
Product Marketing Manager
Nantucket Corporation
12555 West Jefferson Blvd.
Los Angeles, CA  90066
     
---------------------------------------------------------------------
OK, here are the questions:
    
1) First off, to define whether Clipper is a DBMS, we need to clarify
   precisely the definition of what is a DBMS, and what does it do?
        
2) If Clipper is not a DBMS, and it is based mostly on the dBASE
   language, than what makes dBASE a DBMS, and Clipper not?
     
3) Can the programs we create with Clipper be called a DBMS?, Can the
   programs created with dBASE be called a DBMS? - If the answer is
   yes to both, then wouldn't that make Clipper a DBMS?
   
4) Now the hard one....Does a language constitute a DBMS, or do you 
   first have to write a program in that language that let's you
   manipulate the data without programming before it can be called
   a DBMS?
     
I myself do not see a clear definition as to what makes a DBMS and what
does not. Lets see what we come up with here - dLESKO....
 
 
Good Luck....And I hope to see many responses.....


Msg#8090   To:CLIFF GREEN                               (Rcvd)
From:ROBERT HEYMAN                              Date:02/11/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Thanks Cliff


Msg#8091   To:DIRK LESKO                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/11/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
I didn't know that he is with Nantucket...Anyhow I think I remember him
very vaguely...But like Steve(Badaracco said) if we don't solve the
problems here... is less likely we will have them solved by Nantucket.
They have their own...
<CF>


Msg#8092   To:DIRK LESKO                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/11/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8089
------------------------------------------------------------------------------
I think that regardless if Clipper is a DBMS or not(and that for sure
can't be decided by one Publication or by one person(including John
Helliwell who, as a matter of fact I don't know how much he knows about
DBMS)...
Generally speaking, when we we say Clipper seems that we think at a...
compiler, because that's the way it started. Well, know since a lot of
people program in CLipper better that in... a DBMS can we say that it's a
DBMS also ? Probably yes... or probably not...
In spite of positive or negative answer I think neither we or Nantucket
can impose on a Publication like PC Week to include Clipper. PC Week I
think is a private business same like Nantucket and let them run it the
way they want. IF Clipper continues to grow and get most of the
market(even if it is not a BDMS) I bet that it will be in every issue of
PC Week...FREE.
Constantin Florescu


Msg#8093   To:CONSTANTIN FLORESCU                       (Rcvd)
From:JON MELTZER                                Date:02/11/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
yeah, I noticed that ...


Msg#8095   To:WALT BORYS                                (Rcvd)
From:STEVE STRALEY                              Date:02/12/89
Subject:(R)NEW KID IN TOWN AGAIN                Parent Msg#7954
------------------------------------------------------------------------------
Walt, the first two points were/are handled in the ToolkiT and in the
newsletter.  Also, there are some public domain stuff pertaining to the
first issue.  Welcome aboard.
 
Steve


Msg#8096   To:JON MELTZER                               (Rcvd)
From:STEVE STRALEY                              Date:02/12/89
Subject:(R)RTLINK                               Parent Msg#8080
------------------------------------------------------------------------------
I think in the next to last issue of DOSS there was a review of RTLink. 
I don't have it around my desk at the moment, but if anyone out there has
it handy, they can repeat what I said eactly.  I think I gave a barely a
7 out of 10.  FUNcKY got 10..
 
Steve


Msg#8098   To:DIRK LESKO                                (Rcvd)
From:STEVE STRALEY                              Date:02/12/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8089
------------------------------------------------------------------------------
Thanks for the message and I will be contacting them personally.  If I
can stir a little heat, I don't know who else can.  But let me say
something publically.  I believe CLIPPER is more of a DBMS than dBASE! 
And the reasons I say this is simple.  Let's look at the language and at
the very core of it all.  I mean, the point to DBMS is around the
database file and the abilty to work with it.  Clipper can create them
within the language; DBASE needs to have  template file or a dummy file
to start things out unless you're willing to go into the interpreter
environment and do it all by hand.  No, I think this one point should be
brought to light... 
 
Again, these are hasty thoughts on the fly but down deep in the gut,
something tells me that the editors of some magazines are just listening
to a few disgunted people without checking into the facts.  Thanks for
the note Dirk..
 
Steve


Msg#8102   To:DIRK LESKO                                (Rcvd)
From:STEVE DAVIES                               Date:02/12/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8089
------------------------------------------------------------------------------
I talked to John Helliwell for the better part of an hour on the blower
friday and it seems that they had no idea of the amount of loyalty
Nantucket gets from their customer base. I pointed out to him that had 
their buyer's guide had a column for interactive (ie: dot prompt) or 
menu driven interface, then that could have been used as a way of 
pointing out differences and thereby allow Clipper to be included. 
This was the way PC magazine did it last year. One of his points was 
that their readers, Corperate managers and decision makers, would be 
using this type of buyer's guide to decide on their "DBMS", purchase it, 
and then rely on it to do all the wonderful things that a "DBMS" does. 
While this is certainly valid, quite often people in these companies get 
their hands on dBASE III+ for example and spend mega-hours of the 
company's time developing something. As many of us learned the hard way, 
if we had obtained Clipper earlier we would have saved considerable time 
and effort since we re-wrote a lot of that spagetti code anyway (once we 
learned what 'structured code' was).
He told me that they will be putting some sort of little box article
(something like a retraction) in an upcoming issue to let readers know
of the response the've gotten from excluding Clipper from the articles.
He also said they may try to include Clipper in an article in their
'Microscope' section in the near future.


Msg#8105   To:CONSTANTIN FLORESCU                       (Rcvd)
From:KEN CAMP                                   Date:02/12/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8105
------------------------------------------------------------------------------
Constantin,
I'm not sure what you expect to gain by your question, but I deleted my
initial reply, and will try and do the question justice. Curtis Little
and I are co-workers (peers) who work on completely separate programming
tasks. We occasionally work closely, but generally not, however, yes we
do work together. This is fairly common knowledge among the people we met
at the Clipper conference in LA last year, the A-T conference, as well as
on the various CIS Forums we frequent in common. We frequently beta test
one anothers programs, and have on occasion collaborated on programs both
internal within our company (we work for AT&T) and shareware. If someone
feels they were misled, I don't know why. Between the two of us we crank
out a few million lines of code a year. His more math intensive for
analytical purposes, and mine a massive collection of network sutff for
multiple users. I think you'll find a great many of the people on here
have relatively close ties to others........I just don't quite understand
if or why this might be a problem.
 
                        Ken Camp


Msg#8107   To:DAVID MINTER                              (Rcvd)
From:JOHN THORN                                 Date:02/12/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
OK, so much for the California Raisins!  When is Nantucket gonna wise up
and spread these !@#$ conferences around the country!?!?  Let's here it
for having one on the East Coast (maybe the following year in the middle
of the country?).  Anybody else who can't afford the air fare, etc. but
wants to be involved?


Msg#8110   To:GEFF PURCELL                              (Rcvd)
From:RON LANDBERG                               Date:02/13/89
Subject:(R)RECORD CORRUPTION PROB               Parent Msg#8019
------------------------------------------------------------------------------
I realize it was meant in a joking manner   [ "the users are women who
can COMPLAIN LOUDDDDLLYY!!"]  but couldn't you have found some other joke
which was both non-sexist and funny?  It also denigrates the operators of
our customized applications whose jobs are a lot harder than ours and
usually for a lot less money.  After about a week of operating our
wonderful applications, nothing remains for operators, USUALLY, but to do
repetitive, tedious data entry and when problems occur, the frustration
for the operator is every bit as great as for the programmer.
 
As to your problem:  I don't have a solution but I did have a similar
problem for awhile.  Had no explanation for the operators then but
luckily for me it seems to have just gone away.  Good Luck with your
problem.


Msg#8111   To:STEVE BADARACCO                           (Rcvd)
From:JOHN NEWTON                                Date:02/13/89
Subject:(R)SWITCH PROBLEM        
------------------------------------------------------------------------------
Not really.  According to the documentation in the CLIPPER version you
DON'T have to use the drive and memory parameters.  Alsoto prove it to
myself I wrote a small program and used the exact statement and it
worked!  I'm not sure it the problem has someting to do with the way I
layout my procedures and functions and the linking of same.  I hope
Curtis checks in soon and sheds some light on it.  
  
I do appreceiate your "butting" in - feel free to anytime. Its real easy
to miss whats is real obvious to someone else.
 
-John


Msg#8113   To:KEN CAMP                                  (Rcvd)
From:JOHN NEWTON                                Date:02/13/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
I've re-written the program so it is in two aplications and it seems to
work.  I'm not sure if it has something to do with the way I layout my
aplications.  I have all of my procedures and functions - that are call
by one than one module - a seperate file.  This included one that called
SWITCH.  This may or may not have anything to do with the problem I'm
encountering.  BTW, I am using SWITCH22. 
  
As I said I broke the application up into two seprate ones and it seems
to be working.  I just hate having the user to run two seperate programs
when one would be so much better.
  
Any help or ideas would be apreciated.
  
-John Newton


Msg#8114   To:CURTIS LITTLE                             (Rcvd)
From:JOHN NEWTON                                Date:02/13/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
Curtis,
  Thanks.  AS I tols Ken I am trying out the newer version (SWITCH22) and
after breaking the program up it works.  Switch might not like how I've 
linked my application together.  If you could provide a few more
examples, particularly for applications with grouped procedures/functions
and several other called programs it might shead some light on it for me
at least.  I'm one of those folks who learn by examples.
  
I like the potential that SWITCH offers and sure do encourage you to
further refine it (alhought it looks pertty good right now).
 
Thanks, John Newton


Msg#8115   To:CURTIS LITTLE                             (Rcvd)
From:JOHN NEWTON                                Date:02/13/89
Subject:(R)SWITCH!                              Parent Msg#8084
------------------------------------------------------------------------------
Glad to have helped.
 
-John Newton


Msg#8116   To:STEVE STRALEY                             (Rcvd)
From:JOHN NEWTON                                Date:02/13/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8089
------------------------------------------------------------------------------
I'm not sure if we are using the right words or not. DBMS - Data Base
Management System - IS clipper a DBMS or a development system FOR DBMS' ?
  
I use it rather than dBASE and develop programs for end users that meet
their requirements - the finished applications is rather inflexable in
that it is designed for one specific purpose and thats all.  
 
Like Steve, these are rather quick words and just a thought on the
matter.
 
-John Newton


Msg#8117   To:JOHN THORN                                (Rcvd)
From:JOHN NEWTON                                Date:02/13/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
That sounds like a damn good idea to me.
 
-John Newton


Msg#8118   To:KEN CAMP                                  (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/13/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8105
------------------------------------------------------------------------------
Nothing to gain..nothing. Just I like probably some other people wanted
to know...
<CF>


Msg#8121   To:JOHN THORN                                (Rcvd)
From:DIRK LESKO                                 Date:02/13/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8121
------------------------------------------------------------------------------
From initial responses, I beleive PC week is getting the message very
loudly.....I beleive a retraction is imminent.....
 
dLESKO


Msg#8122   To:JOHN NEWTON                               (Rcvd)
From:KEN CAMP                                   Date:02/13/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
I'm not sure I understand the original problem, John. I've been using it
for quite some time without problems, but I did beta test it, and was
familiar with the way it worked. Rather than try to figure it out, I'll
make sure to tell Curtis to call in here when I see him (probably at
coffee break).


Msg#8123   To:JOHN NEWTON                               (Rcvd)
From:CURTIS LITTLE                              Date:02/13/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
John,
  Could you show me what your link line looked like when you had the
program tied together as one?  Switch! should work with it anyway you
want when linked properly.  How close is Switch to the beginning of the
link line?  (If Switch! is at the end it won't work).  For example, say
you have four modules in your program (doesn't matter how many procedures
or functions are in each one).  For simplicity lets say the modules are
called TEST1 TEST2 TEST3 and TEST4.  The following link line wouldn't
work with Switch!:
 
        tlink TEST1 TEST2 TEST3 TEST4 SWITCH87,,,\CLIPPER\CLIPPER ...etc
 
Switch! should be moved right after the first module on the link line as
follows:
 
        tlink TEST1 SWITCH87 TEST2 TEST3 TEST4,,,\CLIPPER\CLIPPER ...etc
 
If there's still a problem when linked with Switch! near the start of the
application please let me know.  After fixing the interface problem in
the older version of Switch! I haven't heard of or found any problems in
the Clipper version of Switch!, so that should get it working.


Msg#8124   To:HARRY GILBERT                              
From:JOE BOOTH                                  Date:02/13/89
Subject:(R)ARRAYS                               Parent Msg#8051
------------------------------------------------------------------------------
Why don't you just do  
  
     FOR x = 1 TO FCOUNT()
        size[x] = STR( size[x],10 )
     NEXT
  
First, before the achoice() call ?


Msg#8126   To:DIRK LESKO                                (Rcvd)
From:JOHN THORN                                 Date:02/13/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8121
------------------------------------------------------------------------------
'Bout time some people get the message!  By the way, Ashton-Tate is STILL
sending me notices to upgrade to the developers release for dB IV.  (Yea,
yea, I have dB3...just adds to my collection of disks that collect dust
and documentation that props the door open on very windy days near the
shore >grin<.  Now with dB IV, you get 14 disks to do nothing with
and even more verbage by the pound!


Msg#8127   To:STEVE STRALEY                             (Rcvd)
From:JON MELTZER                                Date:02/13/89
Subject:(R)RTLINK                               Parent Msg#8080
------------------------------------------------------------------------------
Well, we all know that FUNCkY deserves at least a 15 ...
but, seriously, what problems did you find with RTLink? Does it raise
run-time memory requirements? Is it too slow? Does it incorrectly
link external libraries? Is the user interface lousy?
  I'd really like some feedback on this product ; do you sell
back issues of DOSS?


Msg#8128   To:JOHN THORN                                (Rcvd)
From:JON MELTZER                                Date:02/13/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Not to mention those who can't afford the PRICE ...


Msg#8129   To:JOHN THORN                                (Rcvd)
From:MARK NEIDORFF                              Date:02/13/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
It's not like people have not tried.  We even researched meeting places
in NY and sent a set of recommendations to Nantucket.  We also accounted
for several hundred people who would come to the Developer's Conference
East.  Nantucket wanted nothing to do with it since I haven't heard back
from them.
     <<  Mark  >>


Msg#8130   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/13/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Bob,
        I'll ul a few files I found, plus one I put together.  No
warrantees on any of this stuff (sigh).
        The files are:
        LJMACROS.ARC
        OVERLAY.ARC
        OVERLAYS.ARC
        OVER_LAY.ARC
        CLIPOVLS.ARC
 
        My contribution is the last one, and, truth to tell, I haven't
finished with it:  I wanted to put the sample overlays (*.OVL) into the
database mentioned in the PRG file, and make sure they work (seems
reasonable, no?).  However, too much landed on my desk today for a
reasonable stab at it.  More later.  In fact, if there's enough interest,
I'll try to put together a reasonably functional subroutine, complete
with parameter validation and everything.....  (Uh oh, what'd I just
commit to?)


Msg#8131   To:JOHN THORN                                (Rcvd)
From:STEVE DAVIES                               Date:02/13/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8121
------------------------------------------------------------------------------
Speaking of Ashton-Tate. What I did when I got the offers in the mail
(since they're self addressed/stamped to A-T) was to scribble on them 
"Drop the suit against Fox and I'll consider it" and send it back to them.
I wonder how many others did something like that.


Msg#8132   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/13/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Ooops.  LJMACRO.ARC and OVERLAY.ARC are already here.  Scratch those for
ul from me.


Msg#8134   To:JOHN THORN                                (Rcvd)
From:DIRK LESKO                                 Date:02/13/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8121
------------------------------------------------------------------------------
Know what you mean.....1200 bucks is a lot of money....I know soome
pretty nice cars I could buy.....
 
(used of course)
 
dLESKO


Msg#8135   To:MARK NEIDORFF                             (Rcvd)
From:DIRK LESKO                                 Date:02/13/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
I have spoken with Nantucket about an EAST coast conference myself, and
the general consensus that since it was closer to home (theirs, not ours)
it was easier to send MORE Nantucket types to it. If it were in the East
Coast, wew wouldn't have access to as many of the crew....So in other
words, theyre too cheap to fly the lot out here for a week.....(but if it
were you're company, wouldn't you be too?)
 
dLESKO


Msg#8138   To:DIRK LESKO                                (Rcvd)
From:STEVE STRALEY                              Date:02/14/89
Subject:(R)PLINK86 FOR SUMMER 87 
------------------------------------------------------------------------------
I'm back for a couple of days, but do just come over and tickle the
ivories... just curious about the Silver Paint... I'd like to do a review
in about 4 weeks, so it's no big rush.   Thanks...
 
Steve


Msg#8140   To:JOHN NEWTON                               (Rcvd)
From:STEVE STRALEY                              Date:02/14/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8089
------------------------------------------------------------------------------
I think DBMS means just that as well. Data Base Management System.  Well,
since it can automatically create it's own database's (which happen to be
conveniently dBASE compatable), and since it can manipulate those
files, along with appropriate index, flat, memory, and binary files as
well... and since all of this is internal to the Clipper language
AUTOMATICALLY (which was the argument for the large memory size... the
databsse engine), it seems to me that the article in PC Week was off
base.  I'd like to know if we were to judge computer "journals", if PC
Week would be offended if we refused to acknoledge them.  Again, just a
passing thought on this without any REAL intent or direction...
 
Take care...
 
Steve'


Msg#8141   To:JON MELTZER                               (Rcvd)
From:STEVE STRALEY                              Date:02/14/89
Subject:(R)RTLINK                               Parent Msg#8080
------------------------------------------------------------------------------
I found it slower on my PLINK attempt: i was nesting 4 levels of
overlays.  They do not support internal overlays and with that, then I
must ask what type of external overlays could they truely support if
internal ones are not possible.  There is some work necessary for
converting over the RTLINK system and other than that, in small linking
modules to save time and distribution disks... maybe this is a good
product.
 
As far as back issues, yes we do sell them (actually 4 Seasons) at 4
dollars an issue.  You can call them at 212-599-2141.   Takeo care..
 
Steve


Msg#8145   To:KEN CAMP                                  (Rcvd)
From:JOHN NEWTON                                Date:02/14/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
Curtis dropped me a line along with an evening phone number - thats
support. Actually, by splitting up the program into two seperate
aplications has proved a blessing.  The folks who control the money
wanted to see reports on line for a given grouping, but the connect time
and usage of the main frame were to costly.  So they wanted to do a query
and store the results in a text file, download the file and look at it on
a PC.  The data oly need to be downloaded one a month.  Well, the orginal
requirement caled for the user to automatically call the main frame and
do all of this querying and download the data.  The folks who footed the
bill like my alternative better - they are the only ones who have access
to the main frame and now we don't have to create multiple access for
connecting and downloading.  The bosses get the viewer program and get
their updated data from the worker folk - happey ending!!
  
Now as to what the orginal problem was - I never did figure it out. Oh
well, thus is life.
 
-John


Msg#8146   To:CURTIS LITTLE                             (Rcvd)
From:JOHN NEWTON                                Date:02/14/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
I'll upload an archive file with the entire package in it ..... oops I
may have trashed all of it.  The accounting folks though the two part
system was a super idea! I came out of this one smelling like a rose.
 
I"'ll check and see if I have the orginal and upload it.
 
Thanks, 
        -John


Msg#8147   To:STEVE STRALEY                             (Rcvd)
From:JOHN NEWTON                                Date:02/14/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8089
------------------------------------------------------------------------------
It's funny how one slip up can cause such a riot.  The results may be
rather far reaching.  
  I wonder, is CLIPPER a DML (data manipulation language) or a DDL (Data
defination language) or all of the above - kind of obevious
 
-John


Msg#8148   To:ALL                                        
From:STEVE BADARACCO                            Date:02/14/89
Subject:FILE FORMATS                            Parent Msg#8148
------------------------------------------------------------------------------
Howdy!
Does any of you kind folks have file format(s) for one or more of the
following:
  MULTIMATE ADVANTAGE and ADVANTAGE II
  DISPLAYWRITE 3 and 4
Any of these would be GREATLY appreciated.  Thanxxx.
...Steve


Msg#8149   To:STEVE BADARACCO                           (Rcvd)
From:DIRK LESKO                                 Date:02/14/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
check out a book called popular file formats for the IBM PC, or something
similar, I've seen it at Barnes and Noble
 
dLESKO


Msg#8151   To:JOHN NEWTON                               (Rcvd)
From:KEN CAMP                                   Date:02/14/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
Glad to hear it's sall working for you now, John. I'm sure Curtis will be
glad to assist if you run in to any more problems.


Msg#8152   To:CLIFF GREEN                               (Rcvd)
From:ROBERT HEYMAN                              Date:02/14/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
Cliff!  What a guy!!  will take a look at the files immediatly!
SOMEHOW I have the feeling that I'll have to upgrade the memory in the
laser to do this stuff!  Thanks!!!
Rob


Msg#8153   To:STEVE DAVIES                              (Rcvd)
From:ROBERT HEYMAN                              Date:02/14/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8121
------------------------------------------------------------------------------
Steve... Funny you should mention that.  In a weak moment, I had ordered
the developers release of db IV.  When I though better of it, and called
back to cancel, they asked me why...(grin) The Fox suit was ONE of the
things I mentioned.  Do you think thet care?
Rob


Msg#8154   To:JOHN NEWTON                               (Rcvd)
From:CURTIS LITTLE                              Date:02/14/89
Subject:(R)SWITCH PROBLEM                       Parent Msg#8040
------------------------------------------------------------------------------
If you find the original one I'd be interested in seeing it.  I'd like to
link together the OBJ with a version of Switch! that has debuggin info in
it to see what is wrong.  As long as Switch! isn't the last thing linked
in there shouldn't be a problem.


Msg#8155   To:ALL                                        
From:JAMES MCCORMICK                            Date:02/14/89
Subject:IBM 36                                  Parent Msg#8155
------------------------------------------------------------------------------
Simple Question.  What is an IBM/36 and can it be used to run a 
program that was written in Clipper??
 


Msg#8156   To:STEVE BADARACCO                           (Rcvd)
From:MARK NEIDORFF                              Date:02/14/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
There are two releases to Popular File Formats for the PC.  One or both
should have what you need. 
     <<  Mark  >>


Msg#8157   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/14/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
memory upgrades are nice if your budget allows.  I snuck a 2 meg upgrade
into the purchase of our office laser, when everyone else in the building
was poo-pooing it.  Now, I can ul as many fonts as I want, and any macros
I want.  Now that fonts have proliferated, I've come to feel more than
justified.


Msg#8158   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/14/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
BTW, I just want to make sure you understand that the code in
CLIPOVLS.ARC is untested.  The principles are illustrated, but you'll
want to juggle them around some.  Also, the .OVL files I included are
really meant to be dl'd to the printer by COPY xxxxx/B PRN, not by the
code;  they have the macro ID, initialize, and end codes in them (which,
presumably, your program would do for you and then just ul the macro
itself, surrounded by the housekeeping).


Msg#8159   To:DIRK LESKO                                (Rcvd)
From:JON MELTZER                                Date:02/14/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Maybe we can have a FUNCky developers conference ...


Msg#8160   To:DIRK LESKO                                (Rcvd)
From:JON MELTZER                                Date:02/14/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
The author is Jeff Walden;  publisher Wiley. There are two volumes ...
"File Formats" and "More File Formats".
Contents:  book 1-  Lotus, Ability, dBase III, DIF, MultiMate 3.3,
                    SYLK, IBM Plans+. SuperCalc3, VisiCalc, WordStar
           book 2 - Framework II, Reflex, Rich, SuperCalc4, Super
                    Project Plus, Volkswriter 3, WordPerfect 4.1


Msg#8161   To:JAMES MCCORMICK                           (Rcvd)
From:RICHARD HORWITZ                            Date:02/14/89
Subject:(R)IBM 36                               Parent Msg#8155
------------------------------------------------------------------------------
The System 36 is a mini computer. In my opinion it was a dog out of the
starting gate, but in true IBM fashion, it's functionality had nothing to
do with the fact that MIS directors bought tons of them. Now that LANs
are so popular, I've seen many 36's trying to find homes. 
Sorry to be so long winded, but no, it cannot run clipper applications.


Msg#8162   To:JON MELTZER                               (Rcvd)
From:JOHN THORN                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Yea, the price is another issue that Nantucket needs to deal with.  After
spending big bucks on their product, I hate the notion of paying more big
bucks for support, and even more bi
g bucks on conferences they sponsor just to keep us current.  


Msg#8163   To:MARK NEIDORFF                             (Rcvd)
From:JOHN THORN                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Mark:
 
If memory serves, it was Jan/Feb 88 at a NYMCUG meeting when we all beat
Ray Love about the head and shoulders over this issue.  Also remember it
coming up time and again at ensuing meetings and personal discussions.
I don't want to seem querulous, but it is time to spread the conferences
around the country.  Is Reference(Clipper) "sponsoring" this one, too?
Mmmm, maybe this is why Steve Straley is contracted by another firm for
his training - they go to just about every area of the country!
 
John


Msg#8164   To:STEVE DAVIES                               
From:JOHN THORN                                 Date:02/15/89
Subject:(R)DBMS OR NOT....                      Parent Msg#8121
------------------------------------------------------------------------------
Steve:  The idea of putting the note "Drop the suit against Fox" on the
A-T return envelope is GREAT.  In fact, got another one of their mailings
just the other day....tomorrow's mail will include the aforementioned
"hint".....


Msg#8165   To:DIRK LESKO                                (Rcvd)
From:JOHN THORN                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Dirk:  Ref if it were my company, would I be too cheap to fly people out
to a conference.  It all depends.....if it is nothing more than a "let's
party and talk about how great we are, rah!, rah!" meeting, then no, I
wouldn't send my people to such far flung regions of the universe as
NY.  BUT, if it were a true developers' conference, where concrete issues
were discussed, field reports made, etc.; if I were truly interested in
my installed base, wanted to expand it, wished to increase the product's
media attention, etc, then I would insist on their travel.  I don't see
such meetings as just bitching sessions, but a venue for sharing ideas,
pitfalls, solutions, and alike.  Further, they should more than a method
of displaying wares from others (heard and read some war-stories on last
year's conference).  Any company which refuses to support those who keep
it in business (implied in the refusal to discuss moving the conference
from the West Coast) is missing its mark.  Sure, they encourage develop-
ment of user groups - but these, I believe, are springing up as a
direct result of the lack of Nantucket interest and support to the
developers.  Nantucket must see a return on investment as an incentive
to move the conference around the country.  Creative marketing, media
exposure, statistical analysis of potential attendees, etc., can all be
used to force their hand.  I believe Nantucket has its head in the sand
and is showing its colors on a broken mast.  Their attitude must change
B4(it's)2L8, to paraphrase their own license plate!


Msg#8167   To:STEVE BADARACCO                           (Rcvd)
From:RON LANE                                   Date:02/15/89
Subject:(R)PLINK86 FOR SUMMER 87                Parent Msg#7941
------------------------------------------------------------------------------
Try the line:  OVERLAY CODE  , works for me.


Msg#8168   To:DIRK LESKO                                (Rcvd)
From:JOHN NEWTON                                Date:02/15/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
The name: FILE FORMATS FOR POPULAR PC SOFTWARE,APROGRAMMERS REFERENCE
Author:   Jeff Walden
Publisher:John Wiley & Sons, Inc.
ISBN 471-83671-0


Msg#8170   To:DIRK LESKO                                (Rcvd)
From:STEVE BADARACCO                            Date:02/15/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
Dirk (and Mark) ...
Thank you very much.
I had heard about these books, but could not find them locally; I wasn't
sure if they contained the formats I need, or in which "edition" to find
them.
Will do some more research.
Thanxx again,
...Steve


Msg#8171   To:JON MELTZER                               (Rcvd)
From:STEVE BADARACCO                            Date:02/15/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
I guess this means that none of them has Multimate Advantage and/or
Displaywrite 3,4.
...Steve


Msg#8173   To:JON MELTZER                               (Rcvd)
From:DIRK LESKO                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Boy you guys like to make work for me dont ya....
 
dLESKO


Msg#8174   To:JON MELTZER                               (Rcvd)
From:DIRK LESKO                                 Date:02/15/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
Right, I just say the book uesterday...
 
dLESKO


Msg#8175   To:CLIFF GREEN                               (Rcvd)
From:NOEL MANZANILLA                            Date:02/15/89
Subject:(R)LASER PRINTER         
------------------------------------------------------------------------------
thanks, will take all the advice into account and keep you informed. 
hopefully, I will meet with success and maybe discovering something that
i can sure with you and the rest of the people that have been so kind
with their help. thanks again.
noel


Msg#8176   To:JOHN THORN                                (Rcvd)
From:DIRK LESKO                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
In defense of the price, I see many many ads for seminars that cost about
thhe same, but without nearly the reach. The developers conference in LA
in pretty high class with lots of opportunities. The fee for seminars is
probably reaonable when one considsers the amount of industry people that
you can have direct access to... It's the air fare and Hotel that really
get us East Coasters.....
 
dLESKo


Msg#8177   To:JOHN THORN                                (Rcvd)
From:DIRK LESKO                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
I think I am going to make sure that your message makes it to someones
desk over there....It's a point that shold be heard. Phil Usher will be
coming around these parts to make some special visits to the user groups
so I think they are starting to get the message. I also recieved some
technical info from them (albeit about a year too late) but at least they
are starting to get the message.....
 
dLESKO


Msg#8178   To:ALL                                        
From:NOEL MANZANILLA                            Date:02/15/89
Subject:CORRUPTED RECORDS                       Parent Msg#8178
------------------------------------------------------------------------------
hi, I have had the strangest things going on.  sometimes when i use the
copy command in an application, the copied record picks up some fields
from some other record in the dbf.  this problem happens sporadically
but has happened in three different sites that have the same application
loaded on their pcs.  i recently upgraded the memory of the pc's to 640K
from 512K hoping this would solve the problem.  guess what, it still
happens.  any suggestions?
noel


Msg#8179   To:CLIFF GREEN                               (Rcvd)
From:ANDREW GROSS                               Date:02/15/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
> label cad
 
Unfortunately, this cart is the only one that I know of that can do what
it does, but, like I said, it sure surprised the HP engineers!
 
andrew g.


Msg#8180   To:CONSTANTIN FLORESCU                       (Rcvd)
From:ANDREW GROSS                               Date:02/15/89
Subject:(R)WORKHORSE PRINTER                    Parent Msg#8030
------------------------------------------------------------------------------
Toshiba P321SL
 
Fastest speed: Draft quality @12cpi: 216cps
Slowest speed: Letter quality @10cpi: 60cps
 
andrew g.


Msg#8181   To:RICHARD HORWITZ                           (Rcvd)
From:JAMES MCCORMICK                            Date:02/15/89
Subject:(R)IBM 36                               Parent Msg#8155
------------------------------------------------------------------------------
Thanks, thats what I thought.  I have just come across a customer that
HAS JUST GOTTEN theres and now want it to run the world.  


Msg#8183   To:JOHN THORN                                (Rcvd)
From:STEVE STRALEY                              Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Thought I'd jump in here and let everyone know that Jack Tollefson is no
longer with Reference(Clipper) and has started his own firm called
Reference Pages.  This firm and mine have entered an agreement to go
across the country to provide Clipper training.  This may include Canada,
Germany, and Australia in the not-too-distant futures.  Jack's firm has
also been contract to organize the 1989 developers conference, of which
both I and Neil are on the advisory pannel.  I will make a strong pitch
(almost as strong as the memory issue) that the east coast is getting
tired of being ignored and that sometime before the end of the year,
Nantucket should think about doing a sub-conference in the east coast. 
This does not mean it WILL be in New York.  But at the very least, they
should consider the strong pulling the east coast programmer have on the
way "corporate" america is.  Hopefully, like the memory issue, this horse
will be alive and kicking.
 
Thought I'd just keep y'all informed.  
 
Steve


Msg#8186   To:DIRK LESKO                                (Rcvd)
From:MENACHEM BAZIAN                            Date:02/15/89
Subject:HELP                                    Parent Msg#8186
------------------------------------------------------------------------------
Dirk,
 
I have a few questions for you.
 
#1: I have a clipper application running under DOS. I have to get it
running under UNIX. What would be the best way to do it??
 
It seems to me that there are a few ways t do it. Firstly, I could
translate the code to FoxBase and use FB/Unix. Secondly, I could
translate the code to C with DBX and compile it under Unix.
 
Which leads me to question #2: Are you familiar with DBX? Do you know if
the code it generates will runn IDENTICALLY as the clipper application
WITHOUT CHANGES (I know almost NO C), and is it compilable under UNIX?
 
Thanks.
Menachem


Msg#8187   To:STEVE STRALEY                             (Rcvd)
From:ROBERT HEYMAN                              Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Where then if not NY!!  (sheeeesh)   ( El Capital Del Mundo )


Msg#8188   To:ANDREW GROSS                               
From:CLIFF GREEN                                Date:02/15/89
Subject:(R)LASER PRINTER                        Parent Msg#7918
------------------------------------------------------------------------------
> label cad
 
Shucks.


Msg#8189   To:MENACHEM BAZIAN                           (Rcvd)
From:CLIFF GREEN                                Date:02/15/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
you might want to add to your question (since Dirk seems to have more
access to innards and inner knowledge than most):  When's a UNIX version
of Clipper going to be available?  And on which platforms?  Or, will it
come in ANSI C source (for a hefty fee), and be compilable no matter
which machine you're using (e.g., a MIPS-based box, or an HP slob, or a
386i)?


Msg#8190   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Hey!  How about out here in the Heart of Darkness, away from the bright
City Lights (i.e., the Garden State)?


Msg#8191   To:STEVE STRALEY                             (Rcvd)
From:DIRK LESKO                                 Date:02/15/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Another horse?......Where have you been, I have stopped by a few times
to no avail...
 
dLESKO


Msg#8192   To:ALL                                        
From:CLIFF GREEN                                Date:02/15/89
Subject:DELETED()                               Parent Msg#8192
------------------------------------------------------------------------------
Say, I've been toying around with the use of &alias->(DELETED()) with
mixed results.  Anyone else out there use this formulation as a method of
checking to see if a file has any deletions, without having to go through
and COUNT for them?  COUNTing can be long and tedious if the file has
grown to any size, while &alias->(DELETED()) works very quickly when it
works.  I have to add that there are a few other unexpected results in
this particular application, and I'm not too sure I'm not setting
something somewhere which is screwing me up;  nonetheless, any comments
observations, or war stories will be gratefully read and mulled over.
 
(BTW, I've used "&alias" above for no particular reason.  It's not, in fact
the exact code I'm using in my app, but it gets the idea across of what
I'm trying to accomplish.)


Msg#8193   To:MENACHEM BAZIAN                           (Rcvd)
From:DIRK LESKO                                 Date:02/15/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
DBX won't do the trick, you still need to know alot about C to get the
app working. What about that Xdos compiler? won;t it take your Clipper
produced EXE file and do something with it??? If it can't, then beleive
it or not, I would recommend the FOX for that one....(or learn C)
 
dLESKO


Msg#8194   To:ROBERT HEYMAN                             (Rcvd)
From:STEVE STRALEY                              Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
For purposes of reting space and cost, ironically, Philadelphia is high
on the list.  It's 1.5 hours from NY, a couple more from either Boston or
D.C. and that covers the 4 big areas.  The cost of doing a conference is
NOT going to be nearly as steep as any of the other 3 cities mentioned,
and it is considered a "big" city so it can handle the 1,000 or so
Clipper people.  Who knows, maybe I can see about sponsoring a conference
out here... only if interest is good enough. We'll see..
 
Steve


Msg#8196   To:NOEL MANZANILLA                            
From:RON CALLAHAN                               Date:02/16/89
Subject:(R)CORRUPTED RECORDS                    Parent Msg#8178
------------------------------------------------------------------------------
NOEL:
Are you networked. The only time I've seen corrupted record has been
in network situations where all operating paramaters were not set
correctly. (i.e. FILES= in config.sys etc)


Msg#8198   To:ANDREW GROSS                               
From:CONSTANTIN FLORESCU                        Date:02/16/89
Subject:(R)WORKHORSE PRINTER                    Parent Msg#8030
------------------------------------------------------------------------------
Thanks.
<CF>


Msg#8200   To:ALL                                        
From:LANE WHITE                                 Date:02/16/89
Subject:DATA PROBLEM             
------------------------------------------------------------------------------
Hello everyone, new to this board seems excellent.
I have a problem with a client that someone may be able to help me with.
I have written a small system for this client, he has one computer for a
master computer and a other one in the outer office. The client backups
the data from the master to the other computer every so often using Fast
Back Plus. A problem has arisen. When data is restored to the other
computer it is not the same data that is on the master computer! I have
tried using DOS 3.30 Backup and restore and also COPY with the same
results. I have noticed that with some of the DBf records data in the
memo field is corrupt, full of garbage characters? But some of the
records are fine. Other data from other directorys backup and restore
fine to the other computer. Is dSALVAGE worth investigating? I believe
the DBT file is corrupt dont know how it got that way.
Dirk, FUNCky is a excellent Lib. Also, Netlib, Toolkit and R&R are
excellent products I use.
        Any assistance with this matter will be greatly appreciated!
                Lane.


Msg#8201   To:ALL                                        
From:CLIFF CORCORAN                             Date:02/16/89
Subject:UI VER 2                                Parent Msg#8201
------------------------------------------------------------------------------
I am asking for opinion before plunking down $200 bucks for the
'official' release of the UI programer. I have ver 1 and have used it a
bit for screen building etc. Would very much welcome your thought from
those that beta'ed ver 2.
Thanks Cliff


Msg#8202   To:JAMES MCCORMICK                           (Rcvd)
From:GARY MERINSTEIN                            Date:02/16/89
Subject:(R)IBM 36                               Parent Msg#8155
------------------------------------------------------------------------------
A-T did announce a product which is supposed to allow dBASE code to run
on the SYS 3X series of mini's, although I haven't heard much about it.
You might be able to run the code that way if it is vanilla code.


Msg#8203   To:CLIFF CORCORAN                            (Rcvd)
From:RICHARD HORWITZ                            Date:02/16/89
Subject:(R)UI VER 2                             Parent Msg#8201
------------------------------------------------------------------------------
 I have 2 words about UI2, "BUY IT". If you like UI1 (and I assume you do
since you are considering the $200 for UI2) you know what a great
productivity enhancer it can be (no I don't work for wallsoft). They have
made some substantial improvements on the template lang, and even the
last beta (3) is not really too buggy. I almost never sit down and write
my own code anymore from scratch, UI2 does about 75% of the work for me.
Hope this helps.


Msg#8204   To:DIRK LESKO                                (Rcvd)
From:MENACHEM BAZIAN                            Date:02/16/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
Ouch! The problem as I see it is that the Fox/Unix version that is out is
a version behind DOS foxbase.
 
As to learning C, I would agree with you 100% provided I had the time to
do so.
 
Ah well...Back to the ol' drawin board.
 
>MB<


Msg#8205   To:CLIFF GREEN                               (Rcvd)
From:ROBERT HEYMAN                              Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
awww..you're close enuff!!


Msg#8206   To:CLIFF GREEN                               (Rcvd)
From:ROBERT HEYMAN                              Date:02/16/89
Subject:(R)DELETED()                            Parent Msg#8206
------------------------------------------------------------------------------
What are the unexpected results?


Msg#8208   To:ALL                                        
From:HARRY GILBERT                              Date:02/16/89
Subject:NEW CLIPPERE                            Parent Msg#8208
------------------------------------------------------------------------------
While we're defining whether Clipper is a DBMS, note that Spencer F. Katt
in the Feb 13th issue of PC week says that Nantucket will release this
summer a new release of Clipper which is "object oriented", and based on
Smalltalk. It will be backwards code compatible. Now what is object
oriented? Will my apps look like Macintosh apps, with icons?


Msg#8212   To:DIRK LESKO                                (Rcvd)
From:JOHN THORN                                 Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Dirk:  I have no qualms paying for the conference; I concur that the
air fare is the 
crux of the matter!


Msg#8213   To:DIRK LESKO                                (Rcvd)
From:JOHN THORN                                 Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Dirk:  Thanks for passing it on...software vendors who don't listen to
their users are dead - a.k.a. Micropro, et al.


Msg#8214   To:MARK NEIDORFF                             (Rcvd)
From:JOHN THORN                                 Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8214
------------------------------------------------------------------------------
Mark:  Regarding getting the beasts of the east together....yea, I think
it can be done - maybe a joint effort of the of the over 31 User Groups
located east of the Mississippi, 3d party vendors, and Nantucket.  Now
the sponsor....mmmmm....will think on that one.


Msg#8215   To:STEVE STRALEY                             (Rcvd)
From:JOHN THORN                                 Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Steve:  Knew we could count on ya!  BTW, Brentnall and Morgan did an
article in a vain attempt to quell the memory issue in the Jan/Feb issue
of the Nantucket News....guess they think we're stupid!  THANKS, 
NANTUCKET!


Msg#8217   To:CLIFF GREEN                               (Rcvd)
From:JOHN THORN                                 Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Garden State?  Nahhh, I've got to keep my disks in a sealed environment
to keep them from melting....>grin<


Msg#8218   To:MENACHEM BAZIAN                           (Rcvd)
From:DIRK LESKO                                 Date:02/16/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
I would assume that the suit against SCO is not helping either!


Msg#8220   To:JOHN THORN                                (Rcvd)
From:DIRK LESKO                                 Date:02/16/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Not to mention the RIP-OFF the hotel charges!!
 
dLESKO


Msg#8221   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/16/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
How reliable is it? Are there just certain cases where it doesn't work,
or is it more of a random failure? And (the best question) why does it
work?


Msg#8222   To:HARRY GILBERT                              
From:HOWARD KAPUSTEIN                           Date:02/17/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
There is "Object-Oriented Programming" and "Object-Oriented Design", and
they are NOT the same (I just went through all this in Systems
Programming. Joy...) Basically, OOP is modular coding. C, Modula-2, etc.
allow you to break large tasks (say, an accounting system) into smaller
"modules" or tasks. This is good coding practice since your code is miore
easily portable, modifyable, etc.
   OOD is a totally different way of looking at the problem. For example,
to copy a file, with OOP we would do these steps: 1) Check the files
existence, 2) Read the file, 3) Write the file, and 4) Give a Copy
completed message. With OOD, instead of looking at it from a task point
of view, you look at it as "what data do we have, and how is the data
being manipulated". Kinda like deciding what the data is and what is
being done to it than (under OOP) what we have to. It's a little
compliacted, but it looks like it will become more important in the
future. PErsonally, I can't quite get the hang of it. If you do OOP, to
try to think from the OOD perspective is like looking at the problem from
the exact opposite end. 
   C++ is becoming popular becuase it's basically C, so you get the raw
power of C, but it also has a "higher level of data abstraction", so you
can simulate more abstract data more easily. Basically, you're adding
some data types to C that give you some of the power and relationships of
LISP. Expert systems and AI operations need that kind of abstraction to
do those high-end projects. Sorry I don't remember the text book
definition, but I never quite got the hang of it. If you want, I can pry
out the text and quote you a paragraph or two, if the above isn't helpful
enough.
 
Regarding the comment about Clipper being "object-oriented", that term
has become rather bastardized lately. Do they mean OOP or OOD? Will this
make UDFs more of an integral portion of the language? More importantly,
will this mean the linker can leave out whatever it doesn't need, barring
macro expansion? Which can't susbtitute for commands anyway. My guess is
their development of an OS/2 version is highly flavoring this comment, so
their next version, OS/2 compatible probably, will be influenced by the
DLLs (Dynamic Linked Libraries) of OS/2. (Sidebar: Does anyone know if
Unix uses DLL? I forget where, but some other op system does.) 
Personally, I'm more concerned with their current problems, esp.
regarding memory usage, and if the next release will be delayed to see
what happens with the AT/Fox lawsuit (BTW, does anyone have any current
news, or is it just simmering?) 


Msg#8223   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/17/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
Could you fill  me in on what the current status of the lawsuit involving
A-t, Fox and SCO is? Next wekk I have an interview with SCO, so if I hear
any tidbits I'll drop a msg.


Msg#8224   To:ALL                                        
From:HOWARD KAPUSTEIN                           Date:02/17/89
Subject:WP MERGE FILE                           Parent Msg#8224
------------------------------------------------------------------------------
One last question about WP Merge Files. I have an address set of fields,
with an extra field MailAddr since it has to be international, and some
countries put their zip code, city and country in different orders. Well,
the code is basically
IF && use special address &&
  ? TRIM(MailAddr)
ELSE
  ? City, State Country Zip Code
ENDIF
 
That's not real code, but you get the idea. When it comes time to output
the WP Merge File, I need to know if it's ok to output records with
different numbers of fields. If the special address is used, the record
does not get city, state, country or zip, so it's 3 fields short. Is this
a problem? Just so you know, there's a little UDF I use to check if the
special mailling address is used:
 
FUNCTION UseSpecialAddr
PARAMETERS state
RETURN( IIF("*"$state,.T.,.F.) )
 
If the state has a * in it, use the special address, otherwise use the
normal address. Can WP check the state field in the merge file and pick
which addressing method to use? If so, I'll just dump out all the fields,
and let WP decide. Otherwise, I need more info.


Msg#8227   To:CLIFF CORCORAN                             
From:ANDREW VIOLA                               Date:02/17/89
Subject:(R)UI VER 2                             Parent Msg#8201
------------------------------------------------------------------------------
Cliff,
I have been betaing UI since its first beta release (its up to beta 3). 
The new version iPOWERFUL, but its still very buggy.  The concepts in the
product are very strong.  If you like the idea of writing a template with
a full logical instruction set, then it is for you.  Its worth getting on
board now then paying the big price later.
Andrew


Msg#8228   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/17/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
PM will be running on UNIX...which means there is a good posssiblilty
that an OS/2 PM versioni of Clipper will too....
 
dLESKo


Msg#8229   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/17/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
I have not heard any new news at all....All the stories are the
same.....ED Esber saw FOXPRO at Comdex, stormed out in a Huff, 2 weeks
later sued them, they countered sued saying that AT lifted some
technology when FOX showed them some innards when AT almost bought em
out, and appropriated it for Dbase IV.....If you catch any tidbits, drop
em off here.....
 
dLESKO


Msg#8230   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/17/89
Subject:(R)WP MERGE FILE                        Parent Msg#8224
------------------------------------------------------------------------------
WP 5.0 has some minor programming capbilites like IF.....AND type
baloney. However, I think it would be easier for your Clipper program to
do it.....You cannot change the WP secondary file format once you have
defined it, it will throw WP off....Your best bet is to set a 1 byte
field in your database to .T. if the record is international, then write
an additional record to the WP merge file that contains the ENTIRE
ADDRESS you want to print, Or at least the City, State, Zip and
country...Put this extra field on the end, that way all you need to do is
print the following items....
 
Fname Lname
Title
Cname
Address1
CITYSTATEZIPCOUNTRY
 
The last field has been built by Your program with a little logic like:
 
        if (INTERNATIONAL)
                fwrite(WPmerge,City+" "+COUNTRY+" "+IZIPCODE)
        else
                fwrite(WPmerge,City+" "+State+" "+Zipcode)
        endif
 
Or something similiar....ME, I assign a string to a memvar that looks
like:
 
        WPfield = "Fname+chr(xx)+Lname+chr(xx)+Address+chr(xx)..."
 
        Fwrite(WPmerge,&WPfield)
 
That way I can support a whole bunch of merge formats with the same file
write routine.....It runs a little bit slower, but the versatility to
quickly modify the string (WPfield) lets me change the format of the
Merge file while on line without recompiling!!
 
dLESKO


Msg#8232   To:JOHN THORN                                (Rcvd)
From:MARC SCHNAPP                               Date:02/17/89
Subject:NYMCUG                   
------------------------------------------------------------------------------
To joiin NYMCUG and PADD:
Send $50 to 
PADD 75-22 182nd St Flushing NY 11366
Include day and nite phone #'s and you will get the Linker and other
announcenements, benefits, etc.


Msg#8233   To:JOHN THORN                                (Rcvd)
From:DAVID MINTER                               Date:02/17/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
John, Dirk, I've got a suggestion about the air fare, work for the
Airlines!.  I do, it's probably the only way I get to go to these far
flung places.
 Dave
P.S. not meant to be mean.


Msg#8234   To:JOHN THORN                                (Rcvd)
From:NEIL WEICHER                               Date:02/17/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Here, here!  


Msg#8235   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/17/89
Subject:(R)DELETED()                            Parent Msg#8206
------------------------------------------------------------------------------
>unexpected results with alias->(deleted())
 
Well, it doesn't always return .t. when there are deleted records.  I
thought for a while that it was also not allowing a PACK, but that turned
out to be caused by not being in the correct selected work area
(figures!).  Now, I think I've "worked around" the problem by just moving
into another select area before I issue the function.  Probably should
have done that before anyway.  The upshot is that it's only given me a
problem once since then, and I'm not too sure about that one (I got
pulled away to do some real work at that point, and would have to go back
and recreate the situation).


Msg#8236   To:JOHN THORN                                (Rcvd)
From:CLIFF GREEN                                Date:02/17/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Actually, I was thinking of suggesting a conference in (or around)
Princeton, in the so-called hi-tech corridor, and try to inject a bit of
class  into the admittedly lowbrow Clipper fraternity.


Msg#8237   To:HOWARD KAPUSTEIN                          (Rcvd)
From:CLIFF GREEN                                Date:02/17/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
see my reply above (number?)
It was not reliable at all until I figured I needed to be in another
SELEct area before I issued the call.  Now it's pretty durn reliable. 
Before, it just didn't work because I wasn't watching where I was when I
called it.  Why does it work?  Beats me.  I'd like to test other similar
functions to see if this redirection scam works with them too.  Steve
Straley says that you can use SEEK in a similar fashion from another work
area in a UDF to return (I forget what kind of) a value.


Msg#8238   To:HOWARD KAPUSTEIN                           
From:CLIFF GREEN                                Date:02/17/89
Subject:(R)WP MERGE FILE                        Parent Msg#8224
------------------------------------------------------------------------------
What you can do (in WP5.0) is something like this:
in the primary file, for the field number (let's say its field number 3),
you would have ^F3?^ for the international address field, and ^F4?^ for
the normal address field.
 
Use two address fields.  One for the special address, one for the normal.
 
When your program selects and spits out the data, if UseSpecialAddr()
returns a .T., F3 would have data and F4 would be empty (null), and vice
versa.
 
Then, when WP goes to merge your secondary with the primary, the question
mark in the field number will eliminate the blank line(s) from the empty
field.  It just throws it away.  To quote the manual:  "The question mark
tells WordPerfect to see if the field is blank.  If so, the line,
including the end of line character, is deleted in the merged document."
 
Will that suffice?


Msg#8239   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/17/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
Will let you know if I hear anything new. Thanks.


Msg#8240   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/17/89
Subject:(R)WP MERGE FILE                        Parent Msg#8224
------------------------------------------------------------------------------
I don't have WP, nor have I ever done anything more than a quick peek at
it, so I'm ignorant of its capabilities, but I figured it would be easier
to have Clipper do the proper output than let WP do it. I thought of the
same thing as you, either output the international field or
city+state+country+zip+^R, but I have to check with the higher ups to see
if it's ok. Most likely way to go. Thanks.


Msg#8241   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/17/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
That's pretty interesting. If you're in another work area, have you found
any bugs? Traceable or not. And do you (or anyone else) know if it works
with other functions, and if so, which ones. Would certainly cut down on
some coding, if it's reliable.


Msg#8242   To:STEVE BADARACCO                           (Rcvd)
From:JON MELTZER                                Date:02/17/89
Subject:(R)FILE FORMATS                         Parent Msg#8148
------------------------------------------------------------------------------
Sure looks that way ...


Msg#8244   To:HOWARD KAPUSTEIN                          (Rcvd)
From:ROBERT HEYMAN                              Date:02/18/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Howard....I know the following things work:
alias->(eof())
alias->(found())
and probably a host of others
Rob Heyman


Msg#8245   To:JOHN THORN                                (Rcvd)
From:MARK NEIDORFF                              Date:02/18/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8214
------------------------------------------------------------------------------
"Beasts of the East" is terriffic!   (grin)
     <<  Mark  >>


Msg#8247   To:DIRK LESKO                                (Rcvd)
From:MARK NEIDORFF                              Date:02/18/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
Presentation manager on Unix-
What's going on?  Will PM run as a unix task in a single unix session, or
will it become a graphical front end for unix?  And 1000 other related
questions.
     <<  Mark  >>


Msg#8249   To:ALL                                        
From:GARY PREFONTAINE                           Date:02/18/89
Subject:USING 'C' WITH CLIPPER                  Parent Msg#8249
------------------------------------------------------------------------------
I've just starting writing UDFs in 'C' for my Clipper applications.
It seems like in some instances linking in a 'C' UDF with TLINK will
cause the application to hang - while linking it in with LINK or PLINK
will not. Anyone know a/thing about this? The type of function that seems
to hang involves _retc() and _retclen()(?).
Appreciate any help - BTW, compiler is QuickC 2.0


Msg#8250   To:NOEL MANZANILLA                            
From:GEFF PURCELL                               Date:02/18/89
Subject:(R)CORRUPTED RECORDS                    Parent Msg#8178
------------------------------------------------------------------------------
We had a similar problem,  Make sure you are using the version of the
Summer 87 libs that are dated 12-21-87  2:00a .  This solved our problem.
 One of our programmers here had accidentally gotten the old libs with
out the 2:00a time on them in his Clipper dir.  After we put the new
version in and recompiled everything is working again.
The problem on ours seemed to be after a pack, but it could have also
been happening during copy commands.
Geff Purcell  / David Reynolds


Msg#8251   To:MARK NEIDORFF                             (Rcvd)
From:LANE WHITE                                 Date:02/18/89
Subject:(R)DATA PROBLEM          
------------------------------------------------------------------------------
Hello again, problem seems to have cleared up with the following actions.
I fired up DBU and copied the records to another database in a different
directory on the hard drive. And things seem to backup and restore fine
now. There were still garbage characters in some of the memo fields but
at least the client can update the memo field now. I beleive that the
hard drive is not working properly and suggested that the client have a
tech. person look at it. I have found out from nantucket that I have a
early version of Summer 87 they will ship me updated disks, don't know
why I didnt get them before? I was also using SKIP 0 instead of COMMIT to
write out DBF records. Also Nantcuket told me to use 8 or a multiple of 8
for my buffers statement in config.sys? But the client seems to be back
on track. Thanks Mark.
                        Clipping along, Lane.


Msg#8252   To:MARK NEIDORFF                             (Rcvd)
From:DIRK LESKO                                 Date:02/18/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
I beleive the objective is to make it a front end to UNIX so that
anything written to PM will run unchanged on OS/2 or UNIX...Or at least
with minimal re-writing....
 
dLESKO


Msg#8253   To:GARY PREFONTAINE                          (Rcvd)
From:DIRK LESKO                                 Date:02/18/89
Subject:(R)USING 'C' WITH CLIPPER               Parent Msg#8249
------------------------------------------------------------------------------
There are special considerations when compiling for TC...Also, are you
using the Tlink ver 2.0? it doesn't work....
 
dLESKO


Msg#8255   To:DIRK LESKO                                (Rcvd)
From:JOHN THORN                                 Date:02/18/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
dIRK:  Hotel charges are equally outrageous.  Makes you wonder if someone
is getting a percentage....  Shoot, I usually get to stay at Motel-6
>grin<.
 
Yet another reason to MOVE THE CONFERENCE!!!!


Msg#8257   To:CLIFF GREEN                               (Rcvd)
From:JOHN THORN                                 Date:02/18/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Well, Princeton is OK, and the air IS a little better than northern
Jersey >grin<.  There are plenty of sites which could qualify.  NYC,
Philadelphia, Atlantic City (THAT oughtta appeal to the boys and girls
in Nantucket!), Boston, etc, etc.  Anywhere other than LA for the
umpteenth time!!


Msg#8258   To:DIRK LESKO                                (Rcvd)
From:GARY PREFONTAINE                           Date:02/18/89
Subject:(R)USING 'C' WITH CLIPPER               Parent Msg#8249
------------------------------------------------------------------------------
Dirk - did I say TC ? I meant QuickC Version 2. And I'm still using the
oldest Tlink around.


Msg#8261   To:DIRK LESKO                                (Rcvd)
From:MENACHEM BAZIAN                            Date:02/19/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
Dirk,
 
SOmeone just mentioned about a clipper version for unix might be on the
drawing boards...True??? (Oh PLEASE PLEASE PLEASE)
 
Menachem


Msg#8262   To:ANDREW VIOLA                              (Rcvd)
From:MENACHEM BAZIAN                            Date:02/19/89
Subject:(R)UI VER 2                             Parent Msg#8201
------------------------------------------------------------------------------
Andrew,
 
I have a friend who is betaing UI2 and I have helped him a little bit so
I am somewhat familiar with UI2.
 
I agree that it looks like it can be SUPER. But, according to wallsoft,
they are not ready for release for a few weeks. Also, they are quoting
prices for UI2 at over $500.
 
I am not sure that I can get my boss to go for that on a first run
product.
 
Can you give me some more insight into UI2 and what makes it better than
ui1? I am intimately familiar with UI2 and have some familiarity with
UI2.
 
Menachem


Msg#8264   To:MENACHEM BAZIAN                            
From:DIRK LESKO                                 Date:02/19/89
Subject:(R)HELP                                 Parent Msg#8186
------------------------------------------------------------------------------
I would beleive that a DOS version will be first out...and i f that
works, then maybe a UNIX version. But I still think they'll tie it to
some shell running on top of UNIX...
 
dLESKO


Msg#8265   To:HOWARD KAPUSTEIN                          (Rcvd)
From:CLIFF GREEN                                Date:02/19/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
on very little reflection, the only thing I come up with is that it may
be working from evaluation of some system memory variable(s).  Without
knowing anything about the internals of Clipper, my guess would be that
when the compiler parses a function, it may check to see if any system
variables not normally visible to the naked keyboard are be called or
used.


Msg#8266   To:HOWARD KAPUSTEIN                          (Rcvd)
From:CLIFF GREEN                                Date:02/19/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
scratch or amend my previous answer.  now i'm guessing the -> pointer
sets the search of system variables in motion.  maybe both.  maybe i'm
way off base.


Msg#8267   To:MARK NEIDORFF                             (Rcvd)
From:CLIFF GREEN                                Date:02/19/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
ever see a copy of NeWs running?  or any of the licensed variants?  all
sessions run in either a window, or expanded to fill the screen.  Windows
Uber Alles, from one vendor or another.


Msg#8268   To:JOHN THORN                                (Rcvd)
From:CLIFF GREEN                                Date:02/19/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
OK, I can see going to AC for a conf.  Does registration include credit
at the tables?


Msg#8269   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/19/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
The system variable reference makes sense. Each work area probably has
info (or at least each open file) about recno, found, etc. and if they
were smart, if you say FOUND(), Clipper automatically assumes the
currently selected .dbf and uses its internal variables. IF this is true,
then there should be some way of finding out which variables are stored
for each open file, and their equivalent functions should work fine. I'm
not that aware of Clipper's internals either. Dirk?


Msg#8270   To:CLIFF GREEN                               (Rcvd)
From:JOHN THORN                                 Date:02/19/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8073
------------------------------------------------------------------------------
Ref: AC and table credit.  Sorry, I envision on one-armed bandit credit.
Figure Nantucket has a lock on this already, given the amou
nt they charge for support and the quality of it! >grin<  Donald Trump
would have a field day....mmmm, competetive sailing???


Msg#8271   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/19/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Sort of correct, I always wondered why they never let you do the
following:
 
        select 1
        use <DBF>
 
        select 2
        use <DBF2> 
 
        select 3
        use <DBF3>
 
Now, why not:
 
        if .not. EOF(2)
        
or:
 
        if .not. found(2)
 
Instead of the ALIAS pointer stuff? Al the mechanics are already there.
Each work area has it's own buffer for pointers and EOF flags and the
like. When you SELECT, it gets transfered int the buffer that you can
access with EOF(), found() etc....
 
dLESKO


Msg#8272   To:ALL                                        
From:RON LANDBERG                               Date:02/20/89
Subject:SEEK PROBLEM                            Parent Msg#8272
------------------------------------------------------------------------------
> Problem with SEEK on multi-field key
  
I have a file which is indexed on the concatenation of two fields, both
character.  In one part of the application, a SEEK is done on a variable
that will equal the first part of the key, only.  It works fine.
  
In another part of the application, the same SEEK is done, but this time
within a VALID FUNCtion.  Here, the SEEK on a value limited to the first
half of the key fails.  Interestingly, I reindexed the file on the first
field only, then tried the SEEK again.  This time it worked.  This means
to me that my programming logic is correct - it is the SEEK and INDEX
which is screwing up.
  
Can anyone think of a reason a SEEK on a value that will only match the
first part of a concatenated key works in one part of an application
but fails when attempted within a VALID function ?


Msg#8273   To:HOWARD KAPUSTEIN                          (Rcvd)
From:STEVE BADARACCO                            Date:02/20/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Did I recall that the root of this conversation was to determine HOW MANY
deleted records exist in a work area, by saying alias->(deleted())?????
 
Deleted() only returns the deleted status of the CURRENT RECORD in that
area, and I have not had a problem with it.  If you are trying to get a
total deleted count, you must move the record pointer through the entire
range of records to be checked, and test DELETED() on each record.
 
Am I missing something here?
 
...Steve


Msg#8274   To:DIRK LESKO                                (Rcvd)
From:STEVE BADARACCO                            Date:02/20/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Dirk,
Wouldn't that kill the use of parameters for those UDF's.  As in ...
   alias->( MyUDF(param1,param2,param3) )
 
...Steve


Msg#8275   To:RON LANDBERG                              (Rcvd)
From:STEVE BADARACCO                            Date:02/20/89
Subject:(R)SEEK PROBLEM                         Parent Msg#8272
------------------------------------------------------------------------------
Do you have SET EXACT ON ??
 
...Steve


Msg#8276   To:HARRY GILBERT                              
From:LARRY GOTTLIEB                             Date:02/20/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
I'm new to OOP also -- have looked at it for the last coupe of months or
so.  Having just returned from a conference on object oriented
analysis/design/programming, here are a few thoughts:
 
1.  There is very little agreement as to the definition of "Object
Oriented Analysis".  Object Oriented Design is maturing, but there is
still disagreement on its definition, also.
 
2.  There is still some disagreement about the definition of Object
Oriented Programming.  However, the large majority of the speakers at
this conference seemed to be in agreement that:  OOP replaces the
traditional, procedural model with an "object" -- an association between
procedure and data; procedure calls are treated as messages to this
object to perform a set of actions on its associated data.  The idea
behind all of this is to isolate this data from the outside world.  If
successful, the outside world (i.e., other parts of the program) don't
have to worry if the data structure changes.  There are several buzzwords
that go along with OOP (encapsulation, inheritance, data abstraction).
 
3.  The hope of OOA/D/P is that it will help in the handling of more and
more complex systems.  At the conference, several people from Microsoft
noted the important of object oriented-ness to Microsoft's own
development projects.
 
Confused yet?  I certainly am.  As to whether a language is object
oriented or not, beware of claims by a lnaguage's sponsors.  Perhaps one
of the more honest definitions was arrived at as follows:
 
          Object oriented programming is good.
          Language X is good.
          Therefor, language X is object oriented.
 
Good luck in your investigations.


Msg#8277   To:RON LANDBERG                              (Rcvd)
From:DIRK LESKO                                 Date:02/20/89
Subject:(R)SEEK PROBLEM                         Parent Msg#8272
------------------------------------------------------------------------------
Are you trimming the variable you pass to the valid??? Those extra spaces
could make a difference...
 
I use seek on multi key indexes inside valids and they always work
fine....
 
dLESKO


Msg#8278   To:STEVE BADARACCO                           (Rcvd)
From:DIRK LESKO                                 Date:02/20/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
No. That would stay the same I guess. 
 
dLESKO


Msg#8279   To:LARRY GOTTLIEB                            (Rcvd)
From:DIRK LESKO                                 Date:02/20/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
It seems to me that Object Oriented Design is more of a way of thinking
rather than an actual thing you can buy, touch, feel, see etc. And, once
you've got the thinking down pat, you can apply the 'Objectivenes' To any
language you please. Some languages however will be able to surround your
Object Oriented Design with better structure in the COMPILER end of
it....But for now, all the stuff I'm reading in the Press is confusing.
It appears to me that the writers of the magazines can't agree on what is
an OOP and what is not an OOP......
 
dLESKO


Msg#8280   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/20/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Hopefully in their next release they'll let you specify an optional work
area. I think they didn't (at least one reason why) because if you :
 
SELECT 1
USE file
SELECT 2
CLOSE DATABASES
if eof(1)
...
 
What happens? You're trying to evaluate on a non-existent database. Yes,
I know it's not that hard just to bomb out at that point, and otherwise
it gives us more flexibility, but I've noticed a lot of little
stupidities about Clipper that enable FUNCky and their like to flourish
so well. If anyone from Nantucket is listening, how about mentioning this
when you discuss the next release...


Msg#8281   To:STEVE BADARACCO                           (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/20/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Yes. Supposedly, doing &alias->(deleted()) gives you the same as COUNT
FOR DELETED(), but without actually processing the entire database. Why,
I don't know, it doesn't make sense to me either. As far as I know,
Clipper does not internally track any counter besides the current record
number (RECNO()) and the total number of records (RECCOUNT()). In the
header for a .dbf file the recno() value is stored, and the recno() is a
temporary variable, but I thought neither dBASE, Clipper, Foxbase, or any
of the other compatibles store a deleted count. Am I wrong?


Msg#8282   To:HOWARD KAPUSTEIN                          (Rcvd)
From:ROBERT HEYMAN                              Date:02/20/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Maybe I'm confused here:
Why should &alias->(deleted()) give you a count?  As far as I can see,
this should only yield a .t. or .f.  .  Aren't we just checking the
deleted status of a particular record in another area??  Gee, I might be
way off base here...got to test this.
Rob


Msg#8283   To:HOWARD KAPUSTEIN                          (Rcvd)
From:ROBERT HEYMAN                              Date:02/20/89
Subject:&ALIAS- (DELETED())                     Parent Msg#8283
------------------------------------------------------------------------------
Howard..  just to make sure that I hadn't lost my marbles, ran a little
test as follows:
file1 = "cusmas"
file2 = "venmas"
select 0
use &file1
recall all
go top
do while !eof()
   delete
   skip
   skip
enddo
clear
select 0
use &file2
select &file1
go top
do while !eof()
   select &file2
   ? &file1->(deleted())
   inkey(.2.)
   select &file1
   skip
enddo
.
.
Of course what this does is to deleted every third record in file1
and prints on the screen a series of .T. .F. .F. 's.
I've used this sort of construct for quite some time and have never had a
problem.
.
Rob


Msg#8286   To:CLIFF GREEN                               (Rcvd)
From:MARK NEIDORFF                              Date:02/20/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
Haven't even heard of it.  Will have to take a look when I can find it.
     <<  Mark  >>


Msg#8287   To:LARRY GOTTLIEB                            (Rcvd)
From:RICHARD HORWITZ                            Date:02/20/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
  I was just wondering, with all it's use of objects, is UI Programmer an
OOPS?


Msg#8288   To:STEVE BADARACCO                           (Rcvd)
From:RON LANDBERG                               Date:02/21/89
Subject:(R)SEEK PROBLEM                         Parent Msg#8272
------------------------------------------------------------------------------
No, I don't have SET EXACT ON.


Msg#8289   To:DIRK LESKO                                (Rcvd)
From:RON LANDBERG                               Date:02/21/89
Subject:(R)SEEK PROBLEM                         Parent Msg#8272
------------------------------------------------------------------------------
I know it should work.  I am not trimming the variable passing to the
valid.  But in the test example, the variable was exactly the correct
length of the first part of the key and, in fact, was a known to exist
value.  
 
Usually with this sort of thing (should work, always worked in other
applications, etc.) I find that it is my coding that has some subtle
imperfection that I finally spot.  The only reason I raise this problem
is that the exact SEEK works in another part of the application (same
variable, same coding except that the failing SEEK is inside the Vali
)clause) and that by reindexing the file on just the first field of the
multi-key index, the SEEK then works ok.  
 
But it makes no sense to me to trim the variable. Why would I do that?
The first field in the multi-key is 7 chars.  The variable being passed
is from a GET, initialized to 7 chars, and in the test example, the
variable entered was 7 chars. long.  Wouldn't trimming be a mistake under
these circumstances?
 
Well, I have left the problem behind me - I just permanenetly changed the
key to the single field only.  This didn't cost me too much trouble in
the appolication overall so, until I have a sudden inspiration, I'll
leave it like this.


Msg#8290   To:HOWARD KAPUSTEIN                          (Rcvd)
From:ROB HANNAH                                 Date:02/21/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
No.  &alias->(deleted()) does NOT return a count of deleted records. 
Clipper would have to run throught the entire database in real time to do
that.  IF there was a reserved area in the dbf header for maintaining a
deleted() count, then ok that would work.  Otherwise, no way of knowing
if someone else deleted a record in the interum.
 
alias->(udf()) is exactly the same as:
 
cursel = alias()
select alias
udf()
select select( m->cursel )
 
On a seperate note, the last line above is cleaner than select &cursel -
you don't bomb out if the current select area is empty.
 
ROB


Msg#8291   To:HOWARD KAPUSTEIN                          (Rcvd)
From:STEVE BADARACCO                            Date:02/21/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Howard,
I do not think that alias->(deleted()) will return a count of deleted
records.  The Clipper DOC specifically says that DELETED() returns a
logical value, representing the current record.
 
You can say (delcount()), where delcount() is a UDF along the
lines of ...
   FUNCTION DELCOUNT
   PARAMETERS M_AREA
   PRIVATE OLDAREA,OLDREC
   OLDAREA = SELECT()
   SELECT (M_AREA)
   OLDREC = RECNO()
   COUNT FOR DELETED() TO RETVAL
   GOTO OLDREC
   SELECT (OLDAREA)
   RETURN RETVAL


Msg#8293   To:JOHN THORN                                (Rcvd)
From:CLIFF GREEN                                Date:02/21/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
All the wisecracking aside, I wonder how much real interest and support
there is among the user/developer community for an East Coast conf? 
Strikes me that there's an opportunity for another step in the
'maturation' of this community/market, where conferences and such can
develop in the absence of or at least without the impetus from the sole
source of the product:  Nantucket.  There ought to be enough users groups
and second-party (i.e., addon) developers to be able to put something
together.  So what if Nantucket doesn't want to come to the party?


Msg#8294   To:STEVE BADARACCO                           (Rcvd)
From:CLIFF GREEN                                Date:02/21/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
no, Steve, the question was not whether you could get a count of deleted
records, but rather if you could find out quickly (_without_ going
through the whole file with a COUNT FOR DELETED()) if there were _any_
deleted records.  As it turns out, you can do so with DELETED() passed as
a parameter attached to the alias pointer;  apparently, though, it only
works from another work area.  I, at least, have not been able to make it
work from within the work area I'm interested in - I've had to switch out
and then back into the area of interest.


Msg#8295   To:NEIL WEICHER                              (Rcvd)
From:ROB HANNAH                                 Date:02/21/89
Subject:N_CHECKR()                              Parent Msg#8295
------------------------------------------------------------------------------
Possible bug.  I can't determine if I have a record locked using
n_checkr().  Only if someone else does.  I tried fully expanding it to
n_checkr( recno(), n_stanum() ) to no avail.  Any suggestions as to a
workaround?
 
ROB


Msg#8296   To:ROBERT HEYMAN                              
From:CLIFF GREEN                                Date:02/21/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
hmm.  this is starting to take on a life of its own...  In my initial
comments and question, I pointed out that you should get a .T./.F.
response from alias->(deleted()).  That's what I DO get.  I was wondering
if I was doing something wrong (I got sporadic results - turns out I WAS
doing something wrong), and then if anyone else had played with or knew
anything about this and related phenomena.


Msg#8297   To:ROBERT HEYMAN                             (Rcvd)
From:CLIFF GREEN                                Date:02/21/89
Subject:(R)&ALIAS- (DELETED())                  Parent Msg#8283
------------------------------------------------------------------------------
Ah!  so it does work only on the current record in the other, aliased
dbf?  Stranger and stranger.


Msg#8298   To:MARK NEIDORFF                             (Rcvd)
From:CLIFF GREEN                                Date:02/21/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
NeWs is the networked windowing system Sun developed as their front end
to Unix.  It's been picked up by ("endorsed by") ATT, and I've seen it
running on Suns and Iris machines (Silicon Graphics' hot little boxes). 
Very Slick.  Do see it, to see windows done nicely.


Msg#8301   To:CLIFF GREEN                               (Rcvd)
From:ROB HANNAH                                 Date:02/21/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
You cannot get a count of the number of deleted records without stepping
through the entire database no matter how you try.  Dere b no udder way.
 
ROB


Msg#8302   To:CLIFF GREEN                               (Rcvd)
From:JOHN THORN                                 Date:02/21/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
Cliff:  Agree totally with your premise that the user's groups and
3d party developers COULD do something for the East coast Clipper folks. 
I believe we have to address age-old problems:  who's gonna be in
charge; who's ego will get bent outta shape.  
 
With Ashton-Tate's continual vapor-ware (noting their latest admission
that their "compiler" will be late....memory management problems??)
and the pseudo stuff that Fox does, it looks like Nantucket is in for
the long haul, esp with the PC-Week Kat spilling rumors around and about.
 
OK, so where does PADD sit on this issue?  Where does Steve Straley and
Dirk Lesko sit (among others, like Niel - no slight intended, etc.)?
Are they willing to assist/participate?  Will Nantucket at least give
lip service to the notion and provide first-hand tidbits?  I would 
hazard a guess that most people still see Clipper as nothing more than
a means to make their dBASE code run faster and know little about its
real power (this could be why PC-Week slighted it in a recent review
of DBMSs).  
 
So, how 'bout it folks?  Interested?  I'm willing to help.  Anybody
else?  Suggestions?  Comments?  Are we beating a dead horse?  Is there
any REAL interest?  User groups, how 'bout y'all?  Gurus?  Poor 
consultants trying to make an honest buck and in need of a shot-in-the-
arm?  I once read a message on this board penned by a lady from
Nantucket who proclaimed interest in what was going on.  (I'll resist
the temptation to create the lymric...make up your own.)  Aside from 
that single, short epistle she's grown silent.  Is there a message here?


Msg#8305   To:ROBERT HEYMAN                             (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Oh yes, I agree, it SHOULD give the status of a single record. According
to what I've been told, it acts as a count. Right?


Msg#8306   To:ROB HANNAH                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Oh, so sorry. In the first message it seemed that &a->(f()) would do a
count. I guess I read it wrong. Oops.


Msg#8307   To:STEVE BADARACCO                           (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Yes yes yes. Sorry if I seem a bit testy, but this is the 4th message of
someone telling me I was wrong. I misread the original message. Yes, I
know DELETED() is a simple boolean, each record as a deleted( bit
(sometimes referred to as a tombstone mark by 'real' database peoples)
and DELETED() just tests if that byte is true or false. Very simple
actually. What I was amazed at was in the original message, it seemed
that it &s->(f()) would act as a COUNT, which beats searching the entire
database. How this could be done w/o searching the entire batabase, I
don't know (because the .DBF strucutre does not keep a deleted counter in
the header, only a total (RECCOUNT()) counter ).


Msg#8308   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Let me get this straight. Using s->(DELETED()) is the same as
 
FUNCTION AreDeleted
PRIVATE counter
COUNT FOR DELETED() TO counter
RETURN( IIF(counter==0, .F., .T.) )
 
but the alias method won't search the entire .dbf like the UDF will. Is
this right?


Msg#8309   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Now Now boys, you've let this horse run too far......Seems to me that you
can keep your own count very easily if you had too....You could even
stick in the dbase header if you wanted to...just use fopen() and
fwrite() to stick it in one of those unused bytes......Every time someone
deletes a record, update your own counter. It's your code so you know
whats going on. You could also use fread() to read in large chunks of the
DBF and scan for the "*" asterisk yourself. They will always be at
((FileSize-HeaderSize)/RecSize) offsets.....Piece o' pie s'far s'I can
seeeeee......
 
dLESKO


Msg#8311   To:DIRK LESKO                                (Rcvd)
From:LARRY GOTTLIEB                             Date:02/22/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
You're right.  There are many conflicting claims as to what is or is not
an object oriented programming language.  While OOP can certainly be
practiced TO SOME DEGREE in traditional languages, it is very difficult,
if not impossible, to offer some of the capabilities required (e.g.,
inheritance).  Some languages (for example, Smalltalk) can claim
to be pure OOP languages, in that everything in the language can be
treated as an object.  Others (such as C++) have added OOP features to C,
and offer all OOP features, while still allowing non-OOP practices. 
Still others (such as ADA) have some OOP features, but are missing
others.
 
Inheritance, and the ability to define new data types, seem to have great
potential.  I wonder how Nantucket intends to implement them?


Msg#8312   To:RICHARD HORWITZ                           (Rcvd)
From:LARRY GOTTLIEB                             Date:02/22/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
I confess!  While I have a copy of UI, I have not used it much (yet!).  I
don't think UI can be called an OOP language, but beyond that point it's
too fuzzy for me.


Msg#8313   To:JOHN THORN                                (Rcvd)
From:JOHN NEWTON                                Date:02/22/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
Wel, I for one will dig into my personal pockets to go to such a
conference.  The Army can't/won't send me to the west coast and I sure
can't afford the airfare and hotels myself.  At least on the east coast I
can drive and for the most part hotels can be had for a bit less than on
the west coast.
  
As to helping, I am willing to contribute time and energy to such an
event. So put me down as being willing to work/attend such a conference.
 
                -John Newton


Msg#8314   To:CLIFF GREEN                               (Rcvd)
From:STEVE BADARACCO                            Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Ok, I'll have to check that one out!  (Crow may be on the menu tonight).
 
Still it seems a buggy way for deleted() to perform, because if 
alias->(deleted()) returns .T. if there are ANY deleted records, then it
will be lying to you in the case that the CURRENT record is not deleted.
 
...Steve


Msg#8315   To:HOWARD KAPUSTEIN                          (Rcvd)
From:STEVE BADARACCO                            Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Hey, if you just want your UDF to return .T. if there are ANY
deleted()'s, how about for speed ...
   COUNT FOR DELETED() TO counter WHILE EMPTY(COUNTER)
   ...or...
   COUNT FOR DELETED() TO counter WHILE !DELETED()
...Steve


Msg#8317   To:MENACHEM BAZIAN                            
From:ANDREW VIOLA                               Date:02/22/89
Subject:UI VER 2                 
------------------------------------------------------------------------------
Menachem,
The basic difference is UI2 is a (practically) fully implemented
language.  You can write functions in the UI language that can be called
from within the template, or from other templates.  With UI1, the flow
was top-down, with only a few specific commands (FOR ALL ...) that allows
simply loops.  UI2 has do whiles, cases, and the likes to do elaborate
programs.  For example, I have a template that will make a scrolling
window.  No big deal for UI1 or 2.  But the UI2 template will modify the
generated code to become a valid window (for use with a valid clause)
that optionally pops up the window and modifies the memory variable upon
return, a lookup window thats triggered from a SET KEY to fill in GET
variables, and a normal scroll window. All three can then be optionally
set to be related to a certain key field for one to many relational
lookups.  Try that with UI1, or any other generator.
UI2 is going to be a tremendous program to have if you are a motivated
coder that wants to write templates like this, but will be equally
important it vendors sell templates like this.  Hope that helps, if not
I'll try an explain further.
Andrew


Msg#8318   To:LARRY GOTTLIEB                            (Rcvd)
From:DIRK LESKO                                 Date:02/22/89
Subject:(R)NEW CLIPPERE                         Parent Msg#8208
------------------------------------------------------------------------------
You and me both. If it is really an OOP language, the users are going to
run away with it in as far as defining their own objects and classes. the
support guys are gonna have to learn an awful lot really fast in order to
support the 'Power users' because they are going to be modifying the
behavior of the objects.
 
dLESKO


Msg#8319   To:JOHN NEWTON                               (Rcvd)
From:ROB HANNAH                                 Date:02/22/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
I vote yes for such a beast.  Being a dedicated family man, I don't do
business trips.  But something within a 2hour trip from D.C. would be no
problem.
 
ROB


Msg#8320   To:STEVE STRALEY                              
From:WALT BORYS                                 Date:02/22/89
Subject:(R)NEW KID IN TOWN AGAIN                Parent Msg#7954
------------------------------------------------------------------------------
Steve, thanx for the reply.  Your rev 2 on Clipper is pretty much of
godsend for a (C) programmer trying to get an inside track on inter-
nals (more..more..more, please!). One last shot at packing DBT's;
appreciate the helpful responses from all, but a problem with 
COPY TO..!DELETED() and PACKMEMO.prg.  If there's a major data
base, with limited disk space (just suppose), these strategies
need to copy to a second DBT file, which there may not be room
for.  PACK seems to collapse the DBF in on itself without needing
extra space (at least the Lattice dBC C sources do).  Steve, does
your library routines do this also with the DBT file?
...OTHERWISE, someone mentioned a while ago linked list pointers.
Seems using DBF file with fixed length blank text field and pointers
could be used to emulate a DBT file, with the advantage of being
able to PACK compactly after deletes.  Is it worth writing up
a GETfakememo,PUTfakememo function and uploading it, or am I
reinventing stuff out there?


Msg#8321   To:ROB HANNAH                                (Rcvd)
From:CLIFF GREEN                                Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
>You cannot get a count of the number of deleted records w/out stepping
through the dbf no matter how you try.
 
I know that!  I never asserted any other position!
 
As it turns out, my query about the scope of alias->(deleted()) was
wrong-headed (or at least, the result of an incorrect understanding of
what was going on).  Anyway, I'm still interested in what other functions
work with the "alias->()" combination, and if there are any system memory
variables related to different select areas, etc., etc., etc.


Msg#8322   To:JOHN THORN                                (Rcvd)
From:CLIFF GREEN                                Date:02/22/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
I actually had in mind the users' groups forming a loose federation, with
reps from sufficiently motivated/organized groups coordinating much as
they coordinate their own meetings.  As an example, you'll find national
professional organizations having a National Conference (or Meeting, or
Congress, or whatever pomposity they favor), and Regional Meetings.  The
third party developers could be considered a division of sorts among the
different regions (e.g., the Northeast Clipper Groups as one region, with
developers either being regular members [no reason they can't join a club
too - they're humans too, aren't they?] or having their own association
and representation).
 
Anyway, I think a few reasonably organized and motivated user groups
and/or professional groups (I put PADD in there, to give them the benefit
of the doubt - not being a member myself, and not knowing the real flavor
of the organization) could pull it off.  Certainly industry pundits would
be invited (you need someone to talk to the crowds, about something), and
since some kind of admission would be charged, or fees levied, you could
eventually pay for it.  Hell, you might even be able to tempt someone
from Nantucket to come and speak, if the advance sales were hefty enough.


Msg#8323   To:HOWARD KAPUSTEIN                          (Rcvd)
From:CLIFF GREEN                                Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
That was what I was _hoping_ for, but not what I've found.  I guess this
has been a real red herring.  Mea Culpa. <beat, beat>
 
All s->(DELETED()) does is tell you the boolean attached to the current
record in the area pointed to by s->.  What a letdown.  There oughta be a
way....


Msg#8324   To:DIRK LESKO                                (Rcvd)
From:CLIFF GREEN                                Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Ah!  A constructive suggestion!  Blessed cool water in the desert.
 
Thanks, Dirk.  There had to be a way, and you've just pointed to it.
 
If it's all the same, I'd like to bury this poor horse now.  Unless, of
course someone else has a useful suggestion (keep it clean, now).


Msg#8325   To:STEVE BADARACCO                           (Rcvd)
From:CLIFF GREEN                                Date:02/22/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Yeah, crow can be made delicious...  add lots of garlic.  I've been
munching on this particular scruffy bird for a couple of days, while the
blasted topic refuses to go away.  >sigh<


Msg#8326   To:JOHN THORN                                (Rcvd)
From:RICHARD HORWITZ                            Date:02/22/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
 If you want help setting up the east coast conference, I'm in. Aside
from user groups, I'm sure PLENTY of corporations would send more
programmers to this type of function (no pun intended) if it were held
closer to home. Anyway, let me know what I can do....


Msg#8329   To:CLIFF GREEN                               (Rcvd)
From:JOHN THORN                                 Date:02/23/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8293
------------------------------------------------------------------------------
Cliff:
 
I concur that there needs to be more organization in what groups do.  A
single voice is never as powerful as a full chorus of singers.  Sharing
ideas between groups is presently accomplished via the printed page in
many group newsletters, but nothing beats the eyeball-to-eyeball contact
and resultant question and answer sessions.
 
John


Msg#8330   To:CLIFF GREEN                               (Rcvd)
From:ROB HANNAH                                 Date:02/23/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
alias->() is actually quite useful - not only does it work with udfs, but
you can also phrase "dbf->name + dbf->city + dbf->state" as dbf->( name +
state + city ).  It works with all udfs (SHOULD anyways) and is pretty
much code equivalent to
 
cursel = alias()
select dbf
a = name + city + state
select select( cursel )
 
ROB


Msg#8332   To:CLIFF GREEN                               (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/23/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
I through in just an idea regarding keeping record of the DELETED records
in a DBF. 
What if we would have an auxiliary DBF with only one field like DELETED
numeric or character and every time we delete or add a record we
add/deduct from this field. Do you think that might be possible. This was
just from the top of my head, without to apply my vast
experience(grin)...
<CF>


Msg#8333   To:CLIFF GREEN                               (Rcvd)
From:DIRK LESKO                                 Date:02/23/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
well, someone has to do it....


Msg#8334   To:ALL                                        
From:JAMES MCCORMICK                            Date:02/23/89
Subject:CLIPPER 89 DEV CONFERE                  Parent Msg#8334
------------------------------------------------------------------------------
As far as the Air Fare from NY to LA goes, the rates go 
something like this. Is you stay over a Saturday night 
you can get there and back for about $360.  The problem 
is that Saturday night stay, if your not is a position to 
stay over, the fare can go to over $1000. If I have to pay
my way to the coast, I will book NY to Las Vegas.  The cost
is $220, and they don't care when you come or when you go. 
(Or when you buy your ticket for that matter). Interstate 15 
runs from the Las Veges Strip right into LA. It takes 3-4 hours 
for the trip into the high desert.  Having lived in the Northeast 
most of life, this is a great car ride.  If your not in a rush, 
I recommend Peggy Sues' Dinner outside of Calico, (about half-way 
across), and the Roy Rogers Mus. in Victorville.
(Trigger is there, stuffed.)  As far as 'losing' 4 hours to driving,
the time that I did the trip, I meet someone in Newark who was going 
to the same meeting that I was in LA. (He flew straight in.)  We both 
got to the airport at the same time.  His flight was delayed, and 
we both hit the Hotel at the same time.   So much for the travel-log.  
>JRM<     


Msg#8335   To:ALL                                        
From:MARC SCHNAPP                               Date:02/23/89
Subject:BBS LIST                                Parent Msg#8335
------------------------------------------------------------------------------
I am writing an article for Data Based Advisor on BBS's.
Could you please help me by sharing your favorite BBS 
phone #'s ?  If possible, I would also like the location of
the BBS.  
     
Finally if you REALLY love that BBS, would you be willing
to alert the SYSOP to give me full access so that I can
review the BBS?  Please tell me that you've done so.
      
I will need to complete this project over the next week, so
your immediate response are esssential.           
    
Thanks in advance for your help!!!
    
 


Msg#8337   To:CONSTANTIN FLORESCU                       (Rcvd)
From:CLIFF GREEN                                Date:02/23/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
sure, that's doable (keeping a separate dbf for keeping track of system
variables not kept by Clipper), but I kind of like the compactness of
Dirk's idea - put it in the 'unused' (reserved?) space in the header of
the current dbf.


Msg#8338   To:DIRK LESKO                                (Rcvd)
From:CLIFF GREEN                                Date:02/23/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
yeah, it's a dirty job. ;->


Msg#8339   To:CLIFF GREEN                               (Rcvd)
From:ROBERT HEYMAN                              Date:02/23/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Cliff...  eof()  bof()  found() ...and others I'm sure


Msg#8345   To:DIRK LESKO                                (Rcvd)
From:ALAN GUGGENHEIM                            Date:02/24/89
Subject:SIMPLE WORD PROCESSOR                   Parent Msg#8345
------------------------------------------------------------------------------
     I need to create a simple word processor for notes on patients
in a medical system.  I am concerned about using the memo editor
memoedit() because I have heard that these .dbt files become "bloated"
if you are constantly updating notes.  My questions are the following:
    
1]  Is there really a bloating problem and what can be done?
   
2]  Do you recommend using .txt files instead, and reading them into
    the memoedit() function?
   
3]  THIS QUESTION IS CRITICAL!  WHAT HAPPENS WHEN THE .TXT FILE THAT
    YOU WANT TO READ IN IS TOO LARGE TO FIT INTO MEMORY?!!! 
    IS THERE A WAY OF EDITING LARGE FILES EITHER IN .TXT FORM OR .DBT
    FORM?
  
4]  How can the Funcky lib help me in producing a simple word processor
    using either .txt format or .dbt format?
   
                          Thank You....


Msg#8346   To:STEVE BADARACCO                           (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/24/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
True. Sorry, the gray matter gets grayer as the lights go down (and come
up again...)


Msg#8347   To:CLIFF GREEN                               (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/24/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Oh, I agree, there "ought" to be a way, but take a step back. You're
running under DOS, an operating system lacking any re-entrant
capabilities (thank you MS), using an IBM box (you know, the guys who
figured it would be a great idea to charge you 50% of the "new enhanced
product" with double the speed, when all that was needed was a 5 minute
visit from the service man to move a belt (ah, hearken back to the days
of yesteryear, when IBM's net worth was less than a billion...)), with an
Intel cpu (memory segmentation. Wonderful, aint it?) 
 
I agree, the ought to be an easy way to detect something of such
usefulness, without having to resort to doing it ourselves, but what the
hey. You gets whats you pays for...


Msg#8349   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/24/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Excellent idea. Does anyone know what reserved bytes are used by .dbf
files? The dbf header and field information have reserved bytes, but I'm
not sure which are used and which aren't. Here's the reserved parts:
 
...
DBF header
...
byte reserved1[3];          /* reserved bytes */
byte reserved_multiuser[13];/* reserved for multi-user */
byte reserved2[20];         /* reserved bytes */
 
...
Field structure
...
byte reserved_multiuser1[2];/* reserved for multi-user */
byte reserved_multiuser2[2];/* reserved for multi-user */
byte reserved[8];           /* reserved bytes */
...
 
I presume the reserved_multiuser arrays are used to RLOCK() and FLOCK(),
but does anyone know: 1) exactly how those bytes are set during RLOCK()
and FLOCK(), 2) what other bits are used by dBASE etc., 3) if the
reserved[] arrays are used at all, 4) any other information about other
products that use these bytes.


Msg#8350   To:MARC SCHNAPP                               
From:CONSTANTIN FLORESCU                        Date:02/24/89
Subject:(R)BBS LIST                             Parent Msg#8335
------------------------------------------------------------------------------
Marc, please don't look too far...It's right here. The BOSS is by far
the most prestigious BBS where programers from all USA(I could say even
WORLD) log in everyday. We take pride in having very active on this board
developers like Dirk Lesko, Steve Straley etc.
<CF>


Msg#8351   To:CLIFF GREEN                               (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/24/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
I missed Dirk Sugestion, but if he said something then you forget mine...
Speak to you soon.
<CF>


Msg#8352   To:HARRY GILBERT                              
From:KEVIN CONNEARNEY                           Date:02/24/89
Subject:ARRAYS                   
------------------------------------------------------------------------------
I'VE UPLOADED A FILE CALLED ACHICE2.ARC WHICH WILL ALLOW YOU TO POINT
INTO AND SELECT FROM A NUMERIC ARRAY. iF YOU NEED ANY HELP WITH IT HOLLER


Msg#8353   To:ALAN GUGGENHEIM                           (Rcvd)
From:DIRK LESKO                                 Date:02/24/89
Subject:(R)SIMPLE WORD PROCESSOR                Parent Msg#8345
------------------------------------------------------------------------------
How big are the files you are going to edit? a 60k file holds a lot of
text. you can get at least 15-25 pages in there. I have not seen any real
'Word Processing' type functions other than memoedit(). One solution is
to use these new switchit() or swap() type functions so that you can run
an external editor with no problem. That could let your client use their
familiar Word Processor. FUNCky has no Text editor, it does however have
a bunch of really fast low level file functions that may assist you if
you decide to go with the text file route. If you are storing more than
just a 'memo', you are probably better of creating text files and keeping
track of the filenames in a field in the DBF. It's pretty easy.
 
dLESKO


Msg#8354   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/24/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
I don't have the file structure infront of me, but right after the date
there are 8 bytes that are used for the pointer in memory. Usually there
will be garbage there. I beleive there is some padding right after that,
or you can just create a dummy field that haas a name one character long,
that would give you 10 bytes of buffer to work with yourself. The field
names are in a structure in the file so you always know where they are.
The field name is padded with nulls, or you can just keep overwriting the
name of the field. Clipper will ust think the values stored their is a
field name ... you just ignore it. Make your field 1 char long and it
wont take up too much space...
 
dLESKO


Msg#8355   To:ALL                                        
From:LARRY GOTTLIEB                             Date:02/24/89
Subject:MAX LEVELS OF FUNC CALLS                Parent Msg#8355
------------------------------------------------------------------------------
Does anyone know the maximum number of function/procedure levels allowed
(i.e., a function calls another function calls another function calls
...)?  I've reached 26 levels with no apparent problems, but am worried
about unknowingly trashing memory.
 
If you're wondering how I managed to use so many levels:  a good number
of them involve the use of overlapping windows, employing recursive calls
to ACHOICE, with several of my functions between each call.


Msg#8356   To:MARK NEIDORFF                             (Rcvd)
From:GARY PREFONTAINE                           Date:02/24/89
Subject:(R)USING 'C' WITH CLIPPER
------------------------------------------------------------------------------
Mark, once again thanks. I use the Ram Drive routine - had 3.5 megs of
Ram disk till my AST went bonkers. Now I have 2 Fastram Boards doing
nothing.


Msg#8358   To:LARRY GOTTLIEB                            (Rcvd)
From:ROBERT HEYMAN                              Date:02/24/89
Subject:(R)MAX LEVELS OF FUNC CAL               Parent Msg#8355
------------------------------------------------------------------------------
I was told that the nest level was only 16


Msg#8360   To:JAMES MCCORMICK                            
From:JOHN THORN                                 Date:02/24/89
Subject:(R)CLIPPER 89 DEV CONFERE               Parent Msg#8334
------------------------------------------------------------------------------
Does Peggy Sues' Dinner use what's left of Trigger (stuffed) for
ingredients in the food they serve?


Msg#8361   To:MARC SCHNAPP                               
From:GEORGE SENCOVICI                           Date:02/24/89
Subject:(R)BBS LIST                             Parent Msg#8335
------------------------------------------------------------------------------
Boss is it.  By far!
George Sencovici


Msg#8362   To:ROBERT HEYMAN                             (Rcvd)
From:DIRK LESKO                                 Date:02/25/89
Subject:(R)MAX LEVELS OF FUNC CAL               Parent Msg#8355
------------------------------------------------------------------------------
Nes level when using SET KEY TO....
dLESKO


Msg#8363   To:DIRK LESKO                                (Rcvd)
From:ALAN GUGGENHEIM                            Date:02/25/89
Subject:(R)SIMPLE WORD PROCESSOR                Parent Msg#8345
------------------------------------------------------------------------------
     I am not sure how large these patient notes can get, but I want
this simple word processor to be on a hot-key for simple updating of
a patient's status.  Let's suppose that I use .txt files instead of
memo fields, and want to use memoedit().  What do you suggest I do
in this case, and how can the Funcky library be used?  
     
Suppose these notes are too large to fit into a memory variable for
use by memoedit?  What can be done?  Please refresh my memory on
how the low level functions of FUNCky can help here...
   
                   Thanks...


Msg#8364   To:ALAN GUGGENHEIM                           (Rcvd)
From:MARK NEIDORFF                              Date:02/25/89
Subject:(R)SIMPLE WORD PROCESSOR                Parent Msg#8345
------------------------------------------------------------------------------
Just off of the top of my head, you can use the directory listing
function to find out have big the text file is and then read in only a
part of it using the low level file functions.  Should work well.
     <<  Mark  >>


Msg#8365   To:ALAN GUGGENHEIM                           (Rcvd)
From:DIRK LESKO                                 Date:02/25/89
Subject:(R)SIMPLE WORD PROCESSOR                Parent Msg#8345
------------------------------------------------------------------------------
I would suggest using one ofthe new switchit or swap utilites to run a
small text editor, you'll avoid a lot of headaches and programming...
 
dLESKO


Msg#8366   To:DIRK LESKO                                (Rcvd)
From:ALAN GUGGENHEIM                            Date:02/27/89
Subject:(R)SIMPLE WORD PROCESSOR                Parent Msg#8345
------------------------------------------------------------------------------
     Thanks for the advice!  What you are saying is that making a
word or note processor in Clipper is difficult if you don't know
how large your file to be read in will be.  How do you read in
chunks at at time of the .txt file and read them into the Memoedit()
function, and keep this process looking seamless?
   
     In order to keep my main .exe smaller, I plan to pull out the
memoedit .obj from the extend.lib, and compile it into a separate
overlay.  This technique is useful for any functions in an extended
library that you want to include in an overlay. 
   
Any other comments on using the memoedit() or other functions with
.txt files would be most helpful...


Msg#8367   To:ALAN GUGGENHEIM                            
From:DIRK LESKO                                 Date:02/27/89
Subject:(R)SIMPLE WORD PROCESSOR                Parent Msg#8345
------------------------------------------------------------------------------
You can use fopen() and freadstr() to read in chuncks of the file to
edit, you will have to maintain all your own buffers so that you can put
the file back together when they are done editing. It's not easy, but it
is possible....
 
dLESKO


Msg#8370   To:ALL                                        
From:LANE WHITE                                 Date:02/27/89
Subject:SUBSETS OF DATA                         Parent Msg#8370
------------------------------------------------------------------------------
Hello all! I have a company DBF and part DBF where a company will have
many parts, a one-to-many relationship. Currently I pop up a window to
show part numbers for a given company by loading a array of partnumbers
for only that company, a simple DO WHILE company equals the company
selected after seeking the first company+part number. This seems to work
pretty well but I worry about memory constraints if a company has many
part numbers. Using Dbedit with a filter is very slow. ARe there any
functions out there or is this routine you have to code yourself.
Thoughts are appreciated.
                           Lane
PS. Does anyone know how to call the print screen function from within
clipper. Can you load the keyboard with Prtscr and execute or is this a
function to write in "C" or assembly?


Msg#8371   To:LANE WHITE                                (Rcvd)
From:CONSTANTIN FLORESCU                        Date:02/27/89
Subject:(R)SUBSETS OF DATA                      Parent Msg#8370
------------------------------------------------------------------------------
Lane, I am not sure that I understand exactly the problem on first
question(ref displaying DBF data) but I would try to use m_dbfview or
mx_dbfview(FUNCKy). These functions do what Clipper(and others) probably
will never be able to do. ANd Faaaaaaast ! I have a DBF file with about
15000 records indexed by name, city, phone, product etc and I can look up
in a matter of 1-2 seconds any record, and by any index(name,phone etc)
I am sure that you know Dirk and he will help you with more details.
<CF>


Msg#8372   To:LANE WHITE                                (Rcvd)
From:DIRK LESKO                                 Date:02/27/89
Subject:(R)SUBSETS OF DATA                      Parent Msg#8370
------------------------------------------------------------------------------
To access the print screen routine, just write an assembly routine as
follows:
 
        int     5h
        ret
 
that's all there is to it...Or if you have FUNCky, just execute the
prtscr() function.....
 
dLESKO


Msg#8374   To:DIRK LESKO                                (Rcvd)
From:DIANE LASK                                 Date:02/27/89
Subject:(R)NANTUCKET                            Parent Msg#4836
------------------------------------------------------------------------------
Dirk, I'm so sorry that I've not been able to log on a regular basis. 
Have had modem problems and had to wait till my modem came back.  
I'll be sure to check in on a weekly basis as planned.


Msg#8375   To:MARK NEIDORFF                             (Rcvd)
From:DIANE LASK                                 Date:02/27/89
Subject:(R)NANTUCKET             
------------------------------------------------------------------------------
Hi mark. Sorry it took SOOOO long to get back to you.  Had to get my
modem fixed and it did take a while to get back online.  I'm located in
Los Angeles and work in technical support.  I'll be checking in on a
weekly basis and hope to add any input thayt I can.  Thanks for the
response.
                        Diane.


Msg#8376   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIANE LASK                                 Date:02/27/89
Subject:(R)NANTUCKET                            Parent Msg#4836
------------------------------------------------------------------------------
Hello Howard.  Very sorry to have taken this long to get back to you.
Had a modem crash and took a while to get it back (seems this is apology
week).  I will now be able to check in on a regular basis as planned.
Now to your questions ...
1. To get Turbo C to compile use the following compile line:
                TCC -c -ml -f- -N- -G <proegram name>
and link only with the CL library.  You can't use floating point
operations or screen I/O.  We can't guarantee that everything in Turbo
will work with CLIPPER but I find that most things do.
2.  Optimization of the code linked in is definately one of the things
being strongly looked at in our next release so I really think that you
can expect to see some really good improvement there.
3. U can specify an expression as the breakpoint.  I'm afraid that I'll
have to play around a little with your syntax to see if I can get the
memory expression to work properly.  I'll let u know.
                                                Regards,
                                                Diane


Msg#8377   To:RON LANDBERG                               
From:DIANE LASK                                 Date:02/27/89
Subject:(R)INTRODUCTION                         Parent Msg#7557
------------------------------------------------------------------------------
Hi Ron.  Sorry to take so  long to respond.  Had a modem problem that
kept me offline for a while.  I will be checking in on a weekly basis
from now on.  Well as to the suit,  corporate policy is to let our public
relations people and not the techs respond to such questions.  So if 
you'ld
like to hear from me again I'ld better stick to the policy!  
                                Regards from me and Big Brother,
                                Diane.


Msg#8378   To:DIANE LASK                                 
From:DIRK LESKO                                 Date:02/27/89
Subject:(R)NANTUCKET                            Parent Msg#4836
------------------------------------------------------------------------------
We were wondering where you went to. glad to see that you made it back.
The 'Conferees' were beginning to wonder what happended to you after the
sign on you gave that said you'd be checking in regularily, only to
disappear as swiftly as you appeared. Hope to hear more from you in the
future!
 
dLESKO


Msg#8379   To:DIANE LASK                                 
From:DIRK LESKO                                 Date:02/27/89
Subject:(R)NANTUCKET                            Parent Msg#4836
------------------------------------------------------------------------------
Doesn't Rich's Cgoodies package offer floating point routines for Clipper
that can be used with Turbo C? They're on the board under Cgoodies.arc
 
dLESKO


Msg#8380   To:DIRK LESKO                                (Rcvd)
From:HOWARD KAPUSTEIN                           Date:02/27/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
Where? Here's a header for accessing dBASE files in C:
 
typedef struct DBF {
            byte version;               /* dBASE III version number */
            byte lastupdate[3];         /* date of last update (YY MM DD)
*/
            unsigned long reccount;     /* # records in file */
            unsigned int header_length; /* length of header in bytes */
            unsigned int record_length; /* length of record in bytes */
            byte reserved1[3];          /* reserved bytes */
            byte reserved_multiuser[13];/* reserved for multi-user */
            byte reserved2[20];         /* reserved bytes */
        };
 
Is it here you're talking about? Or in the header for each field (field
name, type, etc.) ? The reserved[] bytes are fine, but what I'm curious
about is if dBASE (or any other product, databases, word proc,
spreadsheets, etc.) use the reserved[] bytes? I presume the
reserved_multiuser[] bytes are for handling FLOCK() and RLOCK() stuff,
but the reserved[] bytes 'should' be unused. I'm curious if anyone
actually messes with these bytes, or is it guarranteed that they are
untouched?


Msg#8381   To:DIANE LASK                                 
From:HOWARD KAPUSTEIN                           Date:02/27/89
Subject:(R)NANTUCKET                            Parent Msg#4836
------------------------------------------------------------------------------
Just a couple more questions, if you know. How about standard stack
frame?I presume line numbers, debug info, etc. should be left out. Also,
I've heard that you can, in general, access unctions like strcpy(), etc.
(MSC functions) via extend, but there are some you have to call with
slightly different names. Does Nantucket have a list of what can be
safely accessed from extend, and what can be accessed if you know what
you're doing, and what should never be accessed? I'm interested in things
like: string manipulation functions (using MSC's functions), Clipper
internal variables (the @...SAY color, @...GET color variables, etc.) and
some other miscellania. Also, any news on the next Clipper version?
Someone mentioned that it will be "more object-oriented", but this
definition is so ambiguous as to be useless. 
   
Thanks. I've finally got some time on my hands so I can afford to see how
well this will work. Will let you know hjow things go. 


Msg#8386   To:ALL                                        
From:HOWARD KAPUSTEIN                           Date:02/27/89
Subject:SUBTREE FIND 3.20        
------------------------------------------------------------------------------
I've just uploaded the latest version of Subtree Find 3.20 (SF320.ZIP)
Here's a brief overview of Subtree Find:
 
    - search any subtree of a disk, from a single directory to the
      entire disk
    - search for files based on file name, attributes, size, date
      and/or time
    - FAST! SF 3.20 outperforms Norton's FF, FA and Graham's Fast [File]
      Finder, by margins of 5% - 250%
    - optionally kill files found, with optional prompting
    - write options to permanently configure SF to your personal tastes
    - optional color output
    - information similar to DIR, but also shows attributes and disk
      space allocated (not just byte size of the file)
    - also search for files inside .ARC files
*** - also search for files inside .ZIP files ***
    - modify attributes (on, off, toggle)
    - 'touch' features allows you to set the date and/or time stamp
      to the current or a specific date and/or time
    - and much more
 
Check out SF320.ZIP today. You won't be disappointed
 
*** = new to SF 3.20
 
SF 3.20 also has several changes and minor bug fixes
 
SF 3.20 requires 53K RAM


Msg#8387   To:CONSTANTIN FLORESCU                       (Rcvd)
From:LANE WHITE                                 Date:02/28/89
Subject:SUBSET DATA                             Parent Msg#8387
------------------------------------------------------------------------------
Hello,
        Thanks for reply. I could use Dbedit and Dirk's function but I
want to pop up a window of data from the DBF based on certain conditions.
Using a filter for a smaller DBF or for a DBF of valid codes works fine
but with large DBF's it is very s l o w... I only want the user to see
part numbers for only that company and no other, I dont want him to
choose the wrong part numbers by showing him all the data. Loading a
array of part numbers for that company and using achoice() for the window
seems to work fine but I worry about memory problems. I guess this is a
design problem with alot of different ways to approach, but the window of
data needs to be fast. I'am looking at conditional indexing but doesn't
seem to be the answer either. Again thanks for your reply.
                                Lane...


Msg#8388   To:DIRK LESKO                                (Rcvd)
From:LANE WHITE                                 Date:02/28/89
Subject:PRINT SCREEN             
------------------------------------------------------------------------------
Thanks Dirk,
        I guess I'am to lazy to look it up in a assembly book!
        Thanks again, Lane...


Msg#8389   To:ALL                                        
From:LANE WHITE                                 Date:02/28/89
Subject:ACCOUNTING PACKAGES                     Parent Msg#8389
------------------------------------------------------------------------------
Hello,
        My background with clients has been in writing special
applications and setting up systems and doing analysis. Lately I have had
clients requesting accounting applications. I have tried to stay away
from accounting applications but I feel I can do this no longer. Does
anyone recommend a accounting application they have worked with? How
about SBT, Sourcemate, Champion, Peachtree ? I'am new to accounting
packages and would appreciate any replys. How about Steve Straleys
accounting package?
                                Lane...


Msg#8390   To:HOWARD KAPUSTEIN                          (Rcvd)
From:DIRK LESKO                                 Date:02/28/89
Subject:(R)DELETED()                            Parent Msg#8192
------------------------------------------------------------------------------
It depends, the header is loaded into memory, and most of the reserved
bytes are used for pointers to memory locations. However, when the file
is closed, the header is re-written, so some of the reserved bytes will
be overwritten with values that have no meaning. If you put something
there, it will be bblown away. A field position can bbe used however. No
one says that a field can't be an identifier for something else....
 
dLESKO


Msg#8392   To:LANE WHITE                                (Rcvd)
From:DIRK LESKO                                 Date:02/28/89
Subject:(R)ACCOUNTING PACKAGES                  Parent Msg#8389
------------------------------------------------------------------------------
I have seen Steves Accounting Package and it looks quite good. I have heard
negative comments about the quality of SBT source code....
 
dLESKO


Msg#8393   To:LANE WHITE                                (Rcvd)
From:CHRISTOPHER EDGAR                          Date:02/28/89
Subject:(R)SUBSET DATA                          Parent Msg#8387
------------------------------------------------------------------------------
Lane,
I had a similar situation and this is what I did: I used the UDF in
dbedit to check for a condition (the field was a key field, so all
records were together in the file) if the condition was evaluated as .t.
the records ws shown on screen.  But, if the condition evaluated to .f. I
had the UDF set the colors to black on black so the record was not shown
and stuffed the keyboard with an up (or down) arrow.  That way you're not
using a SET FILTER TO condition, and you don't have to copy records out
to a temporary file, but you still get to work only with a subset of the
data.  Would this work for you?
 
--Christopher


Msg#8394   To:LANE WHITE                                 
From:CONSTANTIN FLORESCU                        Date:02/28/89
Subject:(R)SUBSET DATA                          Parent Msg#8387
------------------------------------------------------------------------------
Lane, if you want to just display(without editing the data) then you
still can use any one of those two fucntions(m_dbfview or mx_dbfview).
I do this all the time like I said on files as large as 15000 records.
ANd is very fast. Let me know if you want more details.
<CF>


Msg#8395   To:LANE WHITE                                 
From:CONSTANTIN FLORESCU                        Date:02/28/89
Subject:(R)ACCOUNTING PACKAGES                  Parent Msg#8389
------------------------------------------------------------------------------
I recommedn Sourcemate. I like it and you can do anything you like with
the source code. They have an 800 tall free number and are very nice
people. If you call them I am sure they will help you more than other
companies. BTW Even CBS News is using Sourcemate.
PS When you call ask for Mike. You can even use my name. We do some
projects together.
<CF>


Msg#8396   To:ROB HANNAH                                 
From:NEIL WEICHER                               Date:02/28/89
Subject:(R)N_CHECKR()                           Parent Msg#8295
------------------------------------------------------------------------------
No bug.  The Help text on N_CHECKR indicates that the current station is
not included in the list.  However, RLOCK() will return .T. if the record
is already locked at the current station.
 
Would it be useful to have a function that told you if the current
station has the record locked?  Don't you already know that as the result
of the last RLOCK?
 
Neil


