Beijing Paradise BBS Backup/PARADISE.ISO/software/BBSDOORW/LABTST32.ZIP/FIXES.DOC

Text File | 1994-10-15 | 21KB | 480 lines

Labtest 3.2h 10-07-94

As of this date, Labtest is now owned by Omega Software, Imperial Beach,
California.

Added the ability to add a character to each continuation line in the
FILES.BBS file, such as '|'.

I'm targeting the 15th of October as a release date.

Changed the .ANS screen to one drawn by Scott Drake.  An awesome design if
I ever saw one!  Thanks Scott!

Labtest 3.2g 10-04-94

Have I been working on this or what!

Tic processing revisited.  Here's how Labtest will handle the Replaces
command in the .TIC:

Area NET202
Areadesc Net 202 Local Node file distribution area
Origin 1:202/1
From 1:202/707
Magic ECHOLIST
File AREAS.Z62
Replaces AREAS.Z57
Desc NET202 - Combined Net 202 Echo Area listings
Crc 5886CE2B

Labtest will translate the Replaces filename to AREAS.*, thus replacing ANY
AREAS file in the destination directory.  It also removes all occurrences
of the AREAS file from the file data base (for the dest. area).  This is
the way several of us have agreed upon and makes the most sense.  If it's
unpopular to the masses, I'll revisit it again.

The Unregistered/Registered line at the top of the screen doesn't work too
well on RIP screens, so I changed the .RIP file to expand the text window,
this will let it display my little copyright info at the top of the rip
screen as it should.

I'll leave the LTDEBUG environment variable in for user's to use in case of
problems.  The log printing of memory information and command lines will go
away when you remove the LTDEBUG environment variable.  That should trim
the log file somewhat.

In case you're wondering about the delay in starting up when processing
.TIC files, Labtest is setting up some files and stuff.

Labtest will now change the status bar as the echo area changes from .TIC to
.TIC.

My wife and I were talking about something, so I'll let you in on it.  We
plan on publishing a user's manual to give Labtest a professional look.
The manual will be one of those half size jobbers, probably stapled.  I am
open to suggestions on a cover design if any of you are good at that stuff.
This is probably a ways into the future, so don't think it will hold things
up.  Craig Dunstan of TranScan has a printed manual and disk for the
registering users, so I figured I'm going for it.  In case you're
wondering, yes, I am going for the top dog.  I have a feeling with the way
this beta cycle has gone that the users will highly appreciate the amount
of effort we've put into this, and I'm sure it's going to pay off.

The manual will be provided as part of the purchase price.  However, it
will also be available seperately at a reduced cost for anyone wanting to
purchase it.  And of course it'll be free for all beta team members!

Ideas anyone?

Labtest 3.2f 10-03-94

Thanks to Glen Appleton for finding a bug with self extracting archives!

Whew!  That would have been embarassing to release it and find out that it
wouldn't uncompress SFX files.

So, that's fixed.  What else is new?

I added a couple of items to protect my interests.  It would be very easy
for anyone to just have a friend who registered Labtest (or cracked the
code) and just give the key away, since LT doesn't actually say the
registered user's name anywhere except the log file.  You would think that
since Labtest is only $15, it would be worth it to register (IMHO).  But
the cold fact is that with the current methods, if it's so easy to crack
it, why bother registering it right?  So, I've revamped (again) the methods
I use to deter, notice I said deter, not prevent ;), this.  One is that
Labtest now reserves the very first line of the screen for copyright
information. Second is the addition of a LABTEST.KEY file that will be
required to use Labtest in registered mode. Keep in mind there are no
differences between registered and unregistered mode except for the delay
at the beginning.  I hope to some day apply to the ASP, so I cannot (due to
their rules and my ethics), make 50 percent of the features in LT
registered only (like Upcheck).  So what other course could I use?
Comments Please?

Labtest 3.2e 09-30-94

How's this for bug hunting!  I found one when Labtest encounters a file
that has archives lurking in subdirectories.  I was supposed to process
them just as it does anything else.  But NOOOO.  It had to be difficult.
SO, what happened is it just skips them.  Not anymore.  Hope this works on
the OS/2 machines with HPFS (Scott/Hector hint hint)

Major change in the LABTEST.CTL file, instead of using numbers for the
color codes, I'm using actual words to represent the colors.  Much easier
to modify by the sysop wanting to customize his copy.  Still get strange
colors intermittenly on the remote side.  Can anyone else verify this?  But
the new screen (my own design of course) looks a little better I think.

Labtest 3.2d 09-29-94

Fixed a bug with the multinode log file handling.  If you were running on
node 2, Labtest would still write to LABTST1.LOG.  I think this is fixed
now.

Having problems with the 'adjust upload credit'.  How can I properly adjust
the credit when RA is going to credit the user after Labtest is done?
Ideas??

There was a problem with scanning, seems that the second virus scanner
would sometimes get the command line from the first one.  Fixed this.

Added the ability to specify more than one errorlevel for the virus
scanner.  F-PROT can return one of two errorlevels depending one what kind
of virus was detected.  THis way, you can grab both errorlevels and not
have to run the virus scanner twice.  Since I added this, again, the
configuration file has change, so it is recommended that you reconfigure
from scratch, or just double check everything.

Enhanced the scanner log text bit, now instead of just one word, you can
have Labtest search for multiple words or phrases, just separate each
keyword by a space.  THe online help gives more detail.

Added a help screen for the editor pick list window.

Expanded on the adjust user credit help info.

If you didn't specify a filename or path for one of the user messages, a
blank message was written.  This is now fixed.

That's about it for now.

Labtest 3.2c

OOPS!  Skipped this one


Labtest 3.2b

Fixed some problem with the .DIZ importing

Fixed a buglet with the .TIC processing.  It would just stop randomly.  Now
it  will do the whole job as it should.

For some stupid reason Labtest would think a .DOC or other non-archived
file was a .PAK file if it started with a blank line.  This has been fixed.

If a file contained imbedded archives, the file dates were written to the
logfile twice, and the files were also counted twice.  This has been fixed
too.

Labtest 3.2a 09-14-94

Well, not as close as I thought I was to releasing this.

Bugs:

Labtest would not move old .EXE files that were processed (converted from
.EXE to .ZIP ).  Fixed.

Would not check the disk space available on a drive where files were
sent to (mainly during .TIC processing).  This caused a crash, but has been
fixed now.  Sorry Eddy!

I just wanted to acknowledge the immense help Eddy and Miguel Salvador have
been for debugging the .TIC processing code.  Thanks a million!

Added Squish and *.MSG code for ProBoard 2.01 systems.

Labtest 3.2 The final cut!

Finally figured out a fool proof? way of deleting an entry from the file
base even if RAMGR is running.  It works solidly here.  The problem was
that when MOVE was the failed scan action, the old entry was not being
deleted from the file base.  It is now.

Added the virus information writing to the file descriptions.

Supports ProBoard 2.01 and any FILES.BBS system.

Changed LTSetup a bit to move things around.  LTSetup will be going through
a major change in the future, but not before this release is out.

If you do not want Labtest to add your (sysop) name as the uploader in local
mode, you MUST use either the /- switch on the command line or disable Add
Uploader Name in LTSetup.

You can have up to 256 characters in the Sysop messages (text is in
LABTEST.LNG).  You may also use any of the macros that are described in the
documentation for User messages.  If you want to force a carriage return
(to format the sysop's message better), use the ` (reverse single quote) at
the position where you want to start a new line.

The documentation has been split up.  The main user's guide describes
overall setup and use.  The SYSOP.DOC file describes installation into the
BBS software and .TIC file processing in detail.

Added more functionality to the .TIC support.  Look through the SYSOP.DOC
file and the TIC.LAB file for info.

If a file named in a .TIC file already exists in the destination directory,
it will be replaced.  Also, if a .TIC file has the command 'Replaces' in
it, the file being replaced will also be removed.

It's been a long road to this point.  I feel very comfortable with this
release.  It should be a success, and I owe it all to a great beta team! (I
can finally say that!  Never had a beta team before.)

Fixes for Labtest 3.2

Tic Processing revisited.  Various options enabled for the uploader's name
entry.  See TIC.LAB for details.

If there is no room on the destination drive for the incoming file, then
that file will not be processed and its associated .TIC will be renamed to
.BAD.

If the area named in the .TIC file is not configured in your TIC
configuration file, the .TIC file will be renamed to .BAD.

Labtest now checks the free space available on the processing drive.  If it
gets less than that defined in LTSETUP, Labtest will not proceed.  The info
is logged to the log file.

Logs the .TIC file and area being processed.

This will bring me to 3.2a.  The ultimate release will still be named 3.2,
so don't get confused on me...hehehe.

Contemplating some features to process files in the file areas, kind of
like RAMGR.  Ideas?

Fixes for Labtest 3.1j

Fixed two bugs with the TIC processing.  One, it didn't work at all. That's
fixed.  Two, it wouldn't delete the .TIC file when done.  That's fixed too.

Fixed the problem with the MOVE after failure option.  Previously, Labtest
would COPY the file to the badfiles directory, and leave it in the old
directory.  However, it would trash the file's entry in the FDB.HDR file
for the old area.  This has been fixed with one catch.  I noticed today
that under RAMGR, Labtest CANNOT remove the file's record from the header
file for the old area.  This is because of the way Labtest deletes entries,
and I can't think of any better way.  Currently, records are copied from
the .HDR file to a temporary file.  When the bad file's record is found, it
is skipped, and the rest of the records are copied, then the files are
renamed so the temporary file becomes the HDR file.  Problem is RAMGR locks
the file so this prevents renaming it, which causes Labtest to not update
the area file.  If any of you RA guru's have any ideas, please let me know!

Other than this, it moves files just fine, unless you happen to be in
RAMGR!  No problems under RA itself.

Many of you are wondering why it says 0 files uploaded in the thank you
message.  This is because you are running it in LOCAL mode, and in local
mode, Labtest does not read the EXITINFO.BBS file.  This is no problem.  It
works when run online.  Remember, if you are done testing the message
writing portion, remove the LTBETA environment variable!

Any comments on the LABTEST.DOC file?  Good? Bad? Ugly?  Please forward any
left over bug reports as soon as possible.  Would like to get this out
before I reach the letter Z!  haha.

Fixes for LABTEST 3.1I

Minor fixes.  Disabled Online Message option.  Really is unnecessary
especially because RA actually runs Labtest once for each uploaded file.
The user would have several online messages waiting for them.  I only
disabled it so it would be easy to put it back in in the event someone came
up with a good alternative.

During the rewrite of the FDB routines, I leftout the test for drive
letters in the FILES.RA pathnames.  Meaning, if the path did not have a
drive letter, it would not find a match, even though C:\UPLOADS is the same
as \UPLOADS.  This has been fixed.

I finally tracked down the disappearing VIRUS message to the sysop bug.
That was pretty simple after several hours of rewriting and searching.  It
was only a missing carriage return at the end of the message text.  ahhh
the sigh of frustration!

I ran this version on an entire directory and it imported all the file
info, gif resolutions, marked bad files as unlisted, the whole works.  And
even wrote the messages to me and me (user/sysop).

I changed the message to sysop to be the same as message to user, i.e., you
can now use the macros that you use in the user's message in the sysop
message text too.  Just edit your .LNG file and add the macro in the
position you want the message to read.  The enclosed .LNG file has the
sysop's virus message changed to enable appending the VIRUSINFO to the end
of the message.  This way you will get more details than before.

How does Labtest compare with Upcheck/THD/SkullCheck/Etc?  I wanted Labtest
to be on a level of TranScan.  How close am I?

Please send in the straggling bug reports as soon as you can.  People are
beginning to wonder if I'm still supporting Labtest.  And with the bugs in
the current 3.1 that is out, they are getting itchy.  With good reason.


Fixes for LABTEST 3.1H

There was a bug that occurs when you process a virus infected file in your
badfiles directory, the file would be truncated to 0 bytes and the file
information would reflect 0 bytes.  For example, your badfiles area is
C:\BADFILES, and you have the file AVIRUS.ARJ in C:\BADFILES, running
Labtest on this file would mess it up.

Did some work on LTSETUP.  The built-in hard-core limit to the number of
file areas and message areas that LTSETUP can list in the box is 1,000.  If
that is not enough, let me know.

Updated the language file to fix some esthetic things.

I've been asked a billion times about the online messages.  These are
copies of the ones written to the message base.  I know it seems redundant,
but it is actually an alternative to using your message base.  This way,
once the user has read the message, its gone forever.  That could be a
drawback, in case the user wanted to reply to the message, that's why the
message sender is the sysop's name.  Some ideas here would be appreciated.
If all you want to do is tell the user they have mail from Labtest, update
your RA 'Failed Scan' language prompt to read 'File flagged for review.
Please check your mail.' or something like that.

I'm flexible here, so gimme some ideas.

PS.  If any of you do not have a registration key, let me know and I'll
crash it to you.  Its the least I can do for your help, I wish I could do
more.



Fixes for LABTEST 3.1G
LABTEST  EXE     77693 08-25-94   9:28p

I didn't rename this one because no one else got it.

Some more fine tuning with the database routines has killed a couple of
bugs, garbage getting into the header file, and the file comments are
preceded by a carriage return/line feed.  This really doesn't matter but
some sysop's were saying it doesn't get displayed correctly, but they are
probably using the formatted display in RACONFIG, which ignores carriage
returns/line feeds in the descriptions.

Labtest can now be run on a file that is on a CDROM.  When Labtest figures
out that a file is on a cdrom, it sets its defaults to NOREARC, NOCOMMENT
and NOMOVE.  Beware though, when the defaults are changed, they are changed
for the WHOLE session, so mixing cdrom areas with non-cdrom areas won't
work unless you specify the non-cdrom areas FIRST.

The ESC key will abort Labtest, once again!


Fixes for LABTEST 3.1G
LABTEST  EXE     77342 08-24-94   6:00p

Rewrote the file database and message writing routines to try and
enhance the speed and decrease the hard disk use.

Fixed the empty sysop message bug.

Fixed the garbage messages.

Fixed the description problems.

Fixed the Files: xxx Oldest: Newest: where the Newest was almost always
wrong.

The TIC processing hasn't been worked on, but I'd like to keep the feature.
Please help me debug this one too.  I don't want Labtest to stay an
upload scanner its whole life.  Someday, all programs grow up :-)

Some bugs with memory allocation when using Jam have been stomped.

Stack size has been doubled to make life easier when processing entire
directories/file areas.


Fixes for:
LABTEST  EXE     79013 08-17-94   5:28p

Fixed the message writing problem where garbage was getting in the message.

Added the virus scanner program name to the log file info and message info.
Now looks like this:

SCAN.EXE reports:  Found Possessed [Poss] in file DSZ.COM

There was a problem where if the file didn't exist in the database, and it
failed the virus scan, Labtest would not add it to the database.  This only
occurred if you have Failed Scan Action set to MOVE, and the file was
already in your badfiles directory.

Removed the INTERNAL commenting of zip files.  Use the \command.com /c
method for PKZIP.


Fixes for:
LABTEST  EXE     78954 08-15-94   9:16p

Fixed the problem where Labtest would not move a file to the badfiles path
and properly update the file database.    It now does, at least here :(

There was a problem with Labtest not catching the information from SCAN if
you tested an .EXE file and it failed.    Still haven't figured out why it
doesn't remove the @VIRUSINFO@ like it is supposed to.  But still working
on this.  Although Labtest does write the correct info to the message and
log, it just leaves the @VIRUSINFO@ in the message.

Sysop Virus message works here again.

.COM files are treated as .EXE files for virus testing purposes.  The were
treated as UNKNOWN before.

Eddy Salvador reports a problem with Arc Master and the internal comment
routines.  So, if you use AM, use the command.com /c pkzip.exe -z<$COMMENT$
$DIZ$ version instead.    I'll contact Dykstra and find out what's up with
this.

Could someone help me keep track of the bug reports and fixes as I am doing
in this document?  Thanks!

Fixes for:
LABTEST  EXE     78954 08-15-94  12:09a

Gee its late!

Fixed the SECURED -AV zip file problem.  Labtest once again honors the -AV
security.  This broke because I changed the temp filename from CHECKAV.LAB
to CHECKAV.xxx where xxx is the node number, and missed one place where it
still tried to reference the CHECKAV.LAB.  My bust.  Tested this on
GOLFLOG9.ZIP (-AV) and PKZ204G.EXE (-AV SFX).  Both worked.  Labtest
skipped the commenting and repacking like it should.

Eddy Salvador reported a problem with the internal ZIP commenting messing
up ArcMaster (Bill Dykstra).  I can't get AM to view any files on my system
so I can't verify this.  But PKUNZIP -t indicates no problems, the file can
be unpacked just fine and RA's internal archive view works fine.  Help?  I
tried AM on files that had not been processed by Labtest and it wouldn't
show me a directory or anything, just returns back to RA.  The internal ZIP
comment is 99.9 percent effective (allowing .1 percent for CHAOS).  ZIP
comments appear at the end of the zip file and can be located by looking
for the PK\05\06 signature, then seeking 20 bytes forward.  That's it.
Write the length of the comment, then the comment and close the file.
easy. (?)

LABTEST.EXE 08-14-94 6:52p  Fixes

Created my own internal ZIP comment routine.  This only works with ZIP
files.    To use this feature, highly recommended, edit your ZIP command line
and change the PKZIP.EXE -z< to INTERNAL.  Leave the $COMMENT$ $DIZ$.  So
your command should look like this:

    INTERNAL $DIZ$ $COMMENT$ (or whatever your diz/comment is)

Fixed the UNKNOWN and EXE archiving operations.  The files will now be
correctly archived if you have Pack Unknown set to Yes.

Scanning is working.  Errorlevels, log writing and message writing is
working for me.

Sending an online message to the user works great!

   Note that you can have Labtest write a MESSAGE in the message base AND
   send one online, or just write one in the message base.  I haven't
   configured Labtest to ONLY write an ONLINE message and not one in the
   message base.  That's next.