Hi. This is a manual to file/folder scanning
utility that you can use to track changes to the
directory tree on your hard drive.
In this release I did automate many command line options.
Brief explanation of command-line arguments:
To start you type "/Lc:\windows"[command line option] to 
scan c:\windows folder,
or any other drive or folder just as well. If it is a root,
just write "c:", when name has spaces embrace
all in "" astericks. For example you want to scan folder
Program File, type /L"d:\Program Files"
After you press enter, a "dirlog1" ( up to "dirlog7") 
file will be created one for each run of the program on 
your C drive in root. The program watches if there is
already any files in there and makes one with a higher
index. At point where it reaches "dirlog7" the program
deletes all log files and starts clean from "dirlog1"
When you run the option "/compare". It looks in the
same "c:\" folder for the LAST TWO LOG FILES and if finds
run the comparison analysis on them. Output goes into 
"c:\report.txt" file. Only three types
of files are looked for MODIFIED, DELETED, CREATED. I didn't
bother with MOVED because they are basically created files
checked against backed up original folder's files bit by bit with 
new ones(those that are identified as created).
Think of what I just said, it took me a while to figure that
out, you cannot even grasp reading it to say the least.
Now if you have log files somewhere else use this option,
"/F1c:\temp\log1.txt" for the original log or before as I 
call it, and "/F2c:\whatever\log2.txt" for the after file.
Both option separated by a space must be specified 
at the same time to run a comparison. Of course, you
can do it with "/compare" option just as well, by then it picks
last two log files from the root c:\. 

Also make sure there are no very 
long file names. I allocate about 255 chars for a path,
and won't bother to allocate more. It is enough for almost any 
config I've seen so far. 
The compare - text file analysis part is very lousy and slow.
Well, I don't know PERL and C ( with all its barely harnessed
power ) is not the language to search and compare text files.
Unless you set about to write another string library you
won't ever make it with C. Be smart, get somebody's
string library from web and use it.

Also, if you reasonably smart you can cheat this program, at least
you can do that with NT. Trust me on this one, I know something
about Windows platform and FS[file system] driver IO operations.
Normal software would never do whims like that, there simply
would not be a reason for it. However, I know soft folks, most not
guided by logic, etc so they can just code a twist like that.

Also, I have seen some of the techniques I just spoke of in a number
of distinguished products like software antiviruses, back-up, and
fault-tolerance software. It appears what these people learned about
FS and operating system they don't rush to share with us. Neat !
What a bummer!

Now, the scanning part of my algorithm is fast, unlike well-known
recursion method mine is a dynamic function. It is good for several 
reasons, first no need to push data on stack upon each new call to 
the function[itself]. Memory is cheap these days, but the overhead
related to pushing bytes in and out might increase the time of your
program significantly. When we work on a large folder, like WINNT
or WINDOWS or root, something that holds ~10.000 files and folders
is much better worked by a dynamic algorithm than by recursive one.

Analog of this program is FileWatch from cnet.download.com
look at its performance, it is truly awesome, 
especially in comparing mode.

If you like the program or have difficulties running it, please
drop me a line. In a case you are no just a casual reader,
and HAVE TO SEE MY SOURCES you're welcome. Ask me politely,
and you have it.

vortexsny@hotmail.com
Eugene A.
08-11-00