home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Simtel MSDOS 1992 September
/
Simtel20_Sept92.cdr
/
bbs
/
zdcs
/
zdcs_ref.txt
< prev
next >
Wrap
Text File
|
1991-12-31
|
86KB
|
2,112 lines
Zipfile Duplicate Checking System (ZDCS) Ver. 1.65
Copyright (C) 1991, Michael W. Cocke
---------------------------------------------------------------------
Technical Reference Manual
TABLE OF CONTENTS
-----------------
ABOUT THE DOCS..................................................2
PURPOSE OF ZDCS.................................................2
GENERAL TECHNICAL INFORMATION...................................3
ZDCS OPTIONS....................................................4
Deletion of Duplicate Files................................4
Allowed Dupicates..........................................4
BBS Ads....................................................4
Pre-Testing................................................5
INSTALLATION OVERVIEW...........................................5
THE ZDCS CONFIGURATION FILE.....................................5
Function...................................................5
Line 1.....................................................6
Line 2.....................................................6
Line 3.....................................................6
Line 4.....................................................6
Line 5.....................................................7
Line 6.....................................................7
Line 7.....................................................8
Line 8.....................................................8
THE ZDCS DATABASE BUILD.........................................8
Purpose....................................................8
Creation of the Initial Database...........................9
The Screen Display During the Database Build...............9
Additions to an Existing ZDCS Database....................10
ZDCSDB Without a Separate File Integrity Checker..........11
The Database Build Log File ZDCS-DBB.LOG..................11
THE ZDCS DUPLICATE REPORT GENERATOR............................12
THE ZDCS DATABASE PURGE........................................12
BBS ADS........................................................13
Function..................................................13
Creation of the BBS Ads Database..........................14
Updating the BBS Ads Database.............................14
Selecting Deletion or Flagging of BBS Ads.................15
ALLOWED DUPLICATES.............................................16
THE UPLOAD FILE CHECKER........................................17
Purpose...................................................17
Function..................................................17
Maximum and Actual Percent Dupes in an Upload.............18
Updating ZDCS Database(s) After an Upload.................18
Calling ZDCS from EXZTEST.................................19
Calling ZDCSFC from the PCBTEST.BAT file..................19
DOS Error Levels..........................................21
- 1 -
PROCESSING LOCAL UPLOADS.......................................21
PRE-TESTING....................................................22
ZIPs WITHIN ZIPs...............................................24
CD-ROMs........................................................24
AV STAMP INTEGRITY.............................................25
ACCURACY OF THE CRC32 METHOD...................................26
MEMORY LIMITS..................................................26
TROUBLE-SHOOTING GUIDE.........................................26
ZDCSDB (the database build) is crashing...................27
ZDCS is reporting a device I/O error (in any module)......27
ZDCS reports memory corrupt errors (in any module)........27
ZDCS reports path/file access errors......................27
QEMM exception 13 errors occur when running ZDCS..........28
REGISTRATION...................................................29
SUPPORT........................................................30
FUTURE ENHANCEMENTS............................................31
ACKNOWLEDGEMENTS...............................................32
COPYRIGHTS AND LEGAL STUFF.....................................32
ABOUT THE DOCS
--------------
Welcome to the technical reference manual ZDCS-REF.TXT for ZDCS
version 1.65. This manual has been prepared to give you an easy
reference guide for the various parts, functions, and options in ZDCS.
It is deliberately arranged to be modular in organization so that you
can look up whatever interests you without having to read the entire
manual.
There is a second major piece of documentation in this package, a walk-
through called ZDCSWALK.TXT. This is a friendly guide that is meant
to hold your hand and whisper sweet explanations in your ear as you
install and first explore ZDCS. It's arranged to take you through all
the steps you need to know from beginning to end.
PURPOSE OF ZDCS
---------------
ZDCS is a shareware set of utilities intended to help a PCBoard sysop
deal with the problem of duplicate files, whether those files are
already on the bbs or are being uploaded by a caller. It provides
specific support for looking inside ZIP files (including PKZIP version
1.93) and self-extracting files made with PKZIP (SFX). Since both
ZIPs and SFXs are treated the same by ZDCS, we'll refer to both as
zipfiles. ZDCS also provides support for accepting unzipped GIFs.
ZDCS helps to manage the problem of duplicate files in two ways.
1. It provides a method for weeding out duplicate files from an
existing collection of files, like a bbs file system.
2. It provides a duplicate checking method for any prospective
addition to the existing file base, such as uploads.
- 2 -
ZDCS was written by a sysop and programmer who knew that sysops never
have enough time to do all the things they would like to do on their
boards. Making ZDCS friendly to use for both sysop and caller has
been important since version 1.0 first debuted.
Although ZDCS is intended for use on PCBoard bbs's, it can also be
used to look for duplicates on other systems as well, such as a
shareware CD-ROM or even multiple directories on your hard drive
system. However, it's on a bbs that ZDCS really shows its stuff.
ZDCS can be told to decline an upload, to automatically remove
duplicate files, to delete those pesky little bbs ads from uploads,
and to recognize "allowed duplicates" - or any combination of the
above. There's even a pre-test capability that lets callers find out
ahead of time whether or not their intended upload duplicates files
already on your board. All of these are discussed in this reference
manual and described in the walk-through guide ZDCSWALK.TXT as well.
GENERAL TECHNICAL INFORMATION
-----------------------------
ZDCS handles file archives created with the .ZIP extension by PKZIP
(including version PKZIP 1.93) and with the .SFX extension (self-
extracting .EXE files) also created by PKZIP. ZDCS is able to see
inside each of these archives to deal with the individual files inside
them. Since ZIP files are by far more common that SFX files, and
since they are both treated the same way by ZDCS, we'll use the term
"zipfile" to refer to both of them.
ZDCS also handles GIF files. These are individual graphic files that
already have their own compression. GIFs may be uploaded to some
bbs's as individual files. They are not archives of multiple
freestanding files, so ZDCS does not need to look "inside" them. In
fact, ZDCS thinks of them rather like poor zipfiles with only a single
file in them.
Whenever ZDCS encounters a file that is not a ZIP or an SFX, it treats
that file like a GIF. Although we will continue to use the term GIF
to refer to these files, ZDCS could actually handle any "other" type
of file by treating it as a single file, the same as it does with
GIFs.
ZDCS does not provide support for other archiving methods. This means
that if ZDCS encounters an ARJ or LZH archive, for example, it cannot
see inside the archive to look at the individual files. (Support for
other archive formats is under development.) Instead, the entire
archive would be treated as a single GIF-type file.
ZDCS makes use of the 32-bit CRC, often called the CRC32, used
internally by PKZIP for ZIPs and SFXs. When ZDCS encounters any other
file, it considers that file to be a GIF and calculates the CRC32 for
it.
ZDCS checks files for duplicates by keeping a database of the CRC32
values. When new files are added to the system (uploads), ZDCS
- 3 -
compares the CRC32s of the newcomers to those in the database to
determine if there are any matches, indicating duplicate files. This
method uses the identity of the files and is independent of the names
of those files.
The ZDCS database of CRC32s uses a B-tree index, so there are no sort
utilities or regular file maintenance requirements of any kind.
ZDCS is compatible with any Netbios compatible lan, such as Lantastic
or Novell Netware.
ZDCS OPTIONS
------------
ZDCS has four sets of options that can be configured independently of
each other and changed at any time. You can try out the different
options and change your mind about which ones you want to use without
re-installing ZDCS.
For a good guide to choosing the options to fit your system, take a
look at the walk-through ZDCSWALK.TXT. As part of the installation
guide it explains what each option can mean to your bbs and what some
of the possible consequences are. Each of these options is also
covered in greater technical detail in other sections of this
reference manual.
Deletion of Duplicate Files
---------------------------
ZDCS can be set to either flag or delete duplicate files from uploads
to your bbs. Either way, ZDCS will still recognize the files that are
duplicates of ones already in the bbs file system and leaves you
messages in a log file. This deletion feature does not operate on a
GIF file.
Allowed Dupicates
-----------------
You can designate certain files as allowed duplicates. Although these
files may already be present on your bbs, you can tell ZDCS not to
treat them like duplicates. This prevents ZDCS from deleting them if
you have selected the option to delete duplicate files. It also
prevents the allowed duplicates from being counted in the calculation
of the percentage of duplicates in an upload. This feature does work
for GIF files.
BBS Ads
-------
You can specify certain files as truly obnoxious. These are
collectively referred to as bbs ads after one of our favorites. ZDCS
can be told to either flag or delete these files from all uploads.
- 4 -
This option is completely independent of the option to delete
duplicate files. A major distinction is that bbs ads are files that
are known and despised ahead of time. This option provides a safe way
to delete them without risking the removal of authors' unchanged files
from newer shareware versions. The bbs ads option does not operate on
a GIF file.
Pre-Testing
-----------
ZDCS offers a simple method for callers to your bbs to pre-test an
upload before actually sending the full file to the board. The pre-
test is quick and gives the same information that callers would see
after sending the full file. There are no upload credits granted for
using just the pre-test feature without sending up the full file.
INSTALLATION OVERVIEW
---------------------
There are five basic steps to installing ZDCS to work with the bbs:
1. Setting up the configuration file.
2. Creating the initial database.
3. Creating the bbs ads database (optional).
4. Creating the list of allowed duplicates (optional).
5. Setting up the check for uploaded duplicates.
There is a walk-through called ZDCSWALK.TXT included in this package
to help you install ZDCS. It's a friendly guide to understanding the
functions and installing the programs on your bbs. It is especially
good for taking you through the steps slowly and in order so that
installation goes smoothly on your system.
This technical reference manual has information on all the parts,
functions and options of ZDCS, but it's arranged with a different
purpose in mind: to allow you to find and read only those specific
sections you want to know more about. If you want to install ZDCS, we
highly recommend using the walk-through as your primary guide and all-
around hand-holder.
THE ZDCS CONFIGURATION FILE
---------------------------
Function
--------
The ZDCS configuration file ZDCS.CFG is central to the entire process.
All the modules of ZDCS look for and use at least parts of this file.
It's also the place where you set many of the options and all of the
parameters of ZDCS.
For such a powerful component, the ZDCS.CFG is surprisingly simple:
just eight short lines of ASCII that you can create with any text
- 5 -
editor. A sample configuration file is included with this package.
The contents of each line are discussed in excruciating detail further
on in this section.
Under most circumstances, ZDCS.CFG should be in the same directory as
the rest of the ZDCS files. This is always true if you are running
DOS version 3.x or higher.
The one time that the configuration should be located elsewhere is for
those poor souls still running DOS 2.x. In this unusual case,
ZDCS.CFG belongs in whichever directory will be the current directory
when the ZDCS program modules are run.
Line 1
------
This line is the complete drive, path and filename of an ASCII text
file. This is a file that you create listing all the pathnames, one
on each line, that contain the zipfiles / GIFs to be included in the
database. To process a new collection of files into the ZDCS
database, like those on a CD-ROM, just change either this line or the
contents of the file it points to.
There is no upper limit on the number of pathnames that can be
processed. But make sure that you've included the trailing backslash
for each pathname.
If you are not using the index file feature in PCBoard 14.5a, then you
can use the DLPATH.LST file from PCBoard to point to the complete bbs
file system if you want to.
Line 2
------
This line is the drive and pathname where you want the finished ZDCS
database to be located. It makes no difference if you include the
trailing backslash here or not.
You can put the ZDCS database in the same directory as the rest of the
ZDCS files and programs, or you can decide to put it on a different
drive or even across the network.
Line 3
------
This line is either the letter "Y" or the letter "N". It controls
whether you want ZDCS to delete all duplicate files from an upload (Y)
or just flag them and leave them intact (N).
Line 4
------
This line is an integer - that's a whole number, no decimals - between
0 and 100. It sets the maximum percentage of dupes that your bbs will
accept in an upload.
- 6 -
ZDCS will calculate the actual percentage of duplicates in the upload
and compare it to your maximum percentage. If the actual percentage
is lower, the upload is accepted. If the actual percentage is equal
to or higher than the maximum you specified, the upload is declined
and moved to your private upload directory for review.
Setting the percentage to 100 effectively bypasses this filter, since
it permits a duplicated GIF or a zipfile with nothing but duplicates
to pass. At the other extreme, setting the percentage to 0
effectively requires that the uploaded GIFs and zipfiles have no
duplicates at all.
If you make a mistake and enter a decimal on line 4, ZDCS will not
crash. It will simply truncate your number (lop off everything after
the decimal point) and use the resulting integer as the maximum
percentage of dupes.
This works out quite well in actual practice. Uploads that are a
fraction of a percent under the maximum percentage of duplicates are
the only files where this makes any difference, and the use of
truncating instead of rounding means that they will always be passed.
IF ZDCS rounded the decimal points instead, there would be some
uploads that would round up, making their actual percentage of dupes
the same as the maximum value. Such files are declined by ZDCS.
For more discussion of maximum and actual percentage of dupes,
including an example of the above situation, please see the section on
the upload file checker ZDCSFC in this manual.
Line 5
------
This line is the complete drive, path and filename you want ZDCS to
use for the log file created by the upload file checker ZDCSFC. This
log is an ASCII text file that contains information from the upload
file checker ZDCSFC for each upload it has processed. Each message
about a specific upload includes the following:
name of the uploaded zipfile or GIF
list of all component files inside a zipfile
which files are flagged as bbs ads
which files are marked as duplicates
which files are marked as allowed duplicates
actual percentage of dupes in the upload
whether the upload was accepted or declined
If PCBOARD.SYS is in the current directory when the upload file
checker is run, then the name of the currently logged caller is also
included in the log file.
Line 6
------
This line is either the letter "Y" or the letter "N". It controls the
switch to tell ZDCS whether to delete bbs ads (Y) in an uploaded
- 7 -
zipfile or to just flag them (N). If you've decided not to enable any
checking for bbs ads at all, just set this to N.
Line 7
------
This line is reserved for a single line of text by the sysop. The
contents of this line are appended to the PCBFAIL.TXT file whenever an
upload is declined. The caller who has just uploaded the declined
file sees this line of text as a message on the screen.
This line is where you can express from 1 to 72 characters' worth of
creativity. Some callers have become quite fixated on the idea that
"declined" is the same as "thrown out" - which is of course not true.
You can use this line to tell the caller what has happened or will
happpen with the upload. One possible line to use is (without the
quotes) "Too many duplicate files - upload must be reviewed by sysop."
If you don't want to display any message to the caller, just place
something innocuous like a period or even a blank space on this line.
Just don't leave the line completely blank! You also shouldn't use
any quotation marks in this line. You can make use of the PCBoard
@codes; they are fully supported here.
Line 8
------
This line consists of the single letter "Y" or "N". It controls
whether ZDCS displays the one line "registered to" message after the
board receives an upload (Y) or turns off the display of this message
(N). Either way, the caller still sees the file by file breakdown of
the upload and the status (duplicate, bbs ads, etc.) of each file.
This line is only recognized by the registered version of ZDCS. It
has no effect on the three line message displayed by the unregistered
version. It is also the only line of the configuration file that you
can forget to include without causing major problems. If the line is
missing, ZDCS defaults to (Y) and displays the message.
This would probably be a fine time to wax poetical about the
advantages of registration. Instead, we'll just direct you to the
section in this manual on registration for more information about
that, and to the section on support for some of the reasons why.
(Actions speak louder than words.)
THE ZDCS DATABASE BUILD
-----------------------
Purpose
-------
In order for ZDCS to process new uploads, it must first have created a
database for the existing files to serve as a standard of comparison.
- 8 -
This is referred to as the duplicate file database, the ZDCS database,
or the CRC32 database. Information stored in this database includes
the full pathname and file name of every zipfile and GIF examined by
ZDCS and the names and CRC32 values of every individual file.
There is another purpose to building the ZDCS database besides
enabling ZDCS to check uploads. It also serves as a tool to help you
weed out duplicate files from the existing collection.
Creation of the Initial Database
--------------------------------
THE ZDCS database of duplicate files is created with the database
build program ZDCSDB. The finished database consists of three files:
ZDCS.NDX (the index), ZDCS.DAT (part 1 of the data) and ZDCS.PTH (part
2 of the data). A fourth file is also created during this process:
ZDCS-DBB.LOG, which is an ASCII text file logging messages from the
database creation process.
To create the database, you simply run ZDCSDB. This program looks for
the ZDCS configuration file to tell it which files you want to
process. These can be in multiple directories, on different drives,
or across a network. Once you start the build process, there is
nothing more to do until it is complete. Everything will run
automatically with no need for additional input from you. While the
program is running, there is a screen display that gives you the
current status of operations and one very important piece of
information: use the F10 key if you wish to abort the process.
Additional explanation of the screen display is covered in another
section of this manual.
The Screen Display During the Database Build
--------------------------------------------
The ZDCSDB status summary contains both useful and esoteric
information. The most important of these is the notice that the F10
key is the one to use in order to abort the database build. Aborting
the process any other way will almost certainly lead to lost and/or
cross-linked clusters on your hard disk, not to mention a very unhappy
sysop.
In general, ZDCS treats an SFX file like a ZIP file, and this
documentation has reflected that fact by using the term "zipfile" to
refer to both types. For the purposes of this section, we will use
the individual terms SFX and ZIP, since the screen display does
distinguish between them.
One line in the middle of the screen will change frequently as ZDCSDB
is processing files. Besides letting you know that the system is
working, this line tells you something about the processing steps as
they take place, just in case you were curious.
- 9 -
The possibilities are:
Directory C1.. ZDCS has just begun processing a new directory
Directory C2.. ZDCS has just retrieved the next ZIP / SFX / GIF file
within the same directory
ZIP Directry.. ZDCS is retrieving information from within an
individual zipfile
SFX........... ZDCS is retrieving information from within an
individual SFX file
ZIZ........... ZDCS is processing a ZIP within a ZIP
CRC........... ZDCS is calculating the CRC-32 for a GIF file
Indexing...... ZDCS is writing the index and data to the Btree files
Underneath this changing line is a line that tells you which ZIP, SFX
or GIF file is currently being processed, with the full pathname
included. Below that is the phrase "Member files" followed by the
number of individual files within the ZIP, SFX or GIF being processed.
(Remember, a GIF shows up as having only one member file.)
The next line down starts with either "Share" or "NoShare" to indicate
the type of "file opens" being used. The presence of the DOS share
utility in memory is detected by all ZDCS programs to permit automatic
use of the appropriate type of file access.
The second item on the same line is the word "Files" followed by the
number of ZIPs, SFXs and GIFs that have been processed so far.
The third item on the same line is "Members" followed by the total of
all the individual files that have been added to the CRC32 database so
far. Since a GIF is thought of as a ZIP with only one file in it,
each GIF that is processed is counted as just one more individual file
here.
The fourth item is "PDupes" followed by a number. This is arcane
internal status information that will not be explained here.
(Consider yourself lucky.) The only time you are likely to need this
is if there is a probelm with your system during the database build
and we ask you to read the screen display to us.
The very last line is the time of day when the ZDCSDB program was
started. The end time will appear right after the start time on the
same line.
Additions to an Existing ZDCS Database
--------------------------------------
Adding new files to the existing ZDCS database is very simple. If you
run the database build ZDCSDB when there already is an existing ZDCS
database, the records for the new files will be added to those already
in the database. (If you temporarily rename or move the database
files mentioned in the database creation section, then ZDCS will not
recognize them until you restore them.)
This provides a very simple method for adding large or small file
collections to an existing database. All you have to do is edit the
first line of the configuration file to point to the new files, and
- 10 -
you're ready to go. Issue the command ZDCSDB and from here on the
process will proceed just the same way as it does for creating the
initial database.
There is a second way to add files to the ZDCS database. Available
free on The Hacker Central BBS is a merge utility that lets you
combine two ZDCS databases into one. This allows you to create
smaller databases at a time and then merge them together, which can be
an advantage if you have a very large collection to process.
The main advantage of the database merge option is that it allows
sysops to use pre-built ZDCS databases for common file collections,
like the popular CD-ROMs of shareware programs. Pre-built databases
are also available on The Hacker Central and are mentioned in more
detail in the section in this manual on support.
ZDCSDB Without a Separate File Integrity Checker
------------------------------------------------
ZDCS is not a file integrity checker and does not intend to replace
the fine checkers that are already available. Normal operation of
ZDCS assumes that you have already processed your files with a file
integrity checker.
Human nature and circumstances being what they are, a special switch
has been included with the database build module to permit some simple
file integrity checking to be done. This method has ZDCS call in
PKZIP to use the -T switch to test zipfiles. There is no checking
performed on GIFs at all.
To use this feature, you issue the database build command as ZDCSDB T
instead of ZDCSDB. Because ZDCS now calls PKZIP for the file
integrity checking in addition to the usual functions of the database
build, the processing time is greatly increased. If you are only
processing a small collection of files, such as a batch of local
uploads, the difference in time is probably unimportant in practice.
But if you are processing a large set of files, you'll be better off
making your first pass with a file integrity checker and then using
the regular ZDCSDB command to build the database.
If you run ZDCSDB on a collection of files that includes a corrupt
zipfile, the database build will very likely crash. While you can
find out after the fact about the damaged zipfile by reading the log,
it's still far preferable to avoid the problem entirely by using file
integrity checking first.
The Database Build Log File ZDCS-DBB.LOG
----------------------------------------
There is a log file called ZDCS-DBB.LOG created by the database build
operation. This is an ASCII text file that logs messages from the
database creation process.
If ZDCSDB encounters damaged zipfiles while trying to build the
database, there would be messages alerting you to that fact in the log
- 11 -
file. Another case that would generate messages here is when ZDCSDB
hits a zipfile with more than 650 component files. The first 650
files would be processed, and then ZDCS would leave you a message in
the log file letting you know about the situation.
If you have any problems while running ZDCSDB, look in this log file
for help in understanding what happened.
THE ZDCS DUPLICATE REPORT GENERATOR
-----------------------------------
The purpose of the duplicate report generator ZDCSDR is to use the
information in your ZDCS database to give you a report of all
duplicate files in your file collection. The usual time to do this is
after you have created the initial database.
It is entirely likely that when you first create the initial database
you will already have some duplicate files in your collection of
zipfiles and GIFs. To find out about them, run ZDCSDR to generate a
flat ASCII text file called ZDCS-DUP.LST. This is a list of all
duplicate files in the database, including the name and CRC32 of the
duplicated file and the identity with full drive and pathname of the
zipfile or GIF containing the dupe. The format of the ZDCS-DUP.LST
file is the standard comma-separated variable to make it easy to
import this file into a database or parse it into a .BAT file.
When you run ZDCSDR, it asks you whether you want the results sorted
by the CRC32, the individual file name, or the name of the zipfile or
GIF containing the dupe. This sort is called the Wichita sort after
some roundabout and historical reasons.
Once you have the list of duplicates ZDCS-DUP.LST in hand, you are
ready to clean out your file system. Use the information to remove
any duplicate files from the file collection. On some of the larger
bbs file systems, this step alone has freed up megabytes of hard drive
space.
After you are sure you have finished cleaning up the file system, it
is possible to purge duplicate entries from the CRC32 database in
order to reduce its size. This is covered in this manual in the
section on purging the database.
THE ZDCS DATABASE PURGE
-----------------------
Every duplicate file in your initial collection of files on the bbs
will be reflected by a duplicate entry in the CRC32 database, which
ZDCS uses as its standard of comparison for checking new uploads.
Only one entry for each unique file is needed in order to recognize
future uploads of the same file. You can reduce the size of this ZDCS
database by purging duplicate entries from it. This leaves just one
single reference to each individual CRC32 in the database.
- 12 -
Before you make any changes to the ZDCS database at all, please read
the section in this manual on the duplicate report generator ZDCSDR.
The reporting function of this module will supply you with an accurate
list of all your duplicate files, including the necessary path
information - but only if you run it *before* you purge the database.
Without this list, you'll be hard pressed to find and remove the
duplicate files from your system. We strongly recommend that you not
only generate this report but also finish cleaning out the file system
before you purge anything.
When you feel you are ready to purge the database, start by making a
backup copy of it first. This is very important. *Please* make a
copy of the database and put it away somewhere safe before you start
purging info from it. Once you purge this information, the only way
to restore it is to rebuild the database. While it is always possible
to rebuild the database, this takes a fair bit of time, especially for
larger systems. And what sysop has enough time?
By now you are probably really ready to get on with it. To purge the
database of duplicates, you use the duplicate report generator module
ZDCSDR with the P (for Purge) switch. All you have to do is type
ZDCSDR P and hit the enter key. The "P" will redirect the program to
purging the duplicates instead of creating a report on them. There is
nothing else to do until the purge is complete. This is not a fast
process, but it will pay you back with a smaller database.
A second way of modifying the ZDCS database is to use the database
editor, a free utility available on The Hacker Central BBS. This is
actually a debugging tool, but it also works to modify the database
and to delete specific records. Unlike the database purge with ZDCSDR
P, you can't let this one run on its own with no input from you. On
the other hand, the database editor give you greater flexibility over
which duplicate entries you delete.
ZDCS doesn't quite have the intelligence to recognize a bbs ad unless
you tell it which files to look for. To do this, you have to create
the bbs ads database.
BBS ADS
-------
Function
--------
ZDCS offers a way to recognize specific annoying files that can be
targeted for deletion or simply flagged for manual removal,
independently of whether or not you have configured ZDCS to delete
duplicate files. The distinction here is that the files, known
collectively as bbs ads, are *known* files that you have given ZDCS
specific orders to deal with.
In order to recognize a file as a bbs ad, you must first tell ZDCS
about it. This is done by setting up the bbs ads database. After
that, every time ZDCS seems the same file again it knows that it's a
bbs ad.
- 13 -
This option is completely independent of the option to delete
duplicate files, so you don't have to take a chance on removing
authors' unchanged files from newer shareware versions. The bbs ads
option does not operate on a GIF file.
Creation of the BBS Ads Database
--------------------------------
All you need to create the bbs ads database is the program ZDCSBA and
a collection of bbs ads, pyramid schemes, chain letters or other
obnoxious files.
Once created, the bbs ads database will consist of a single file, ZDCS-
BBA.NDX, which will be located in the same directory as the rest of
the ZDCS files. You can delete this file at any time if you want to
start over again with a fresh collection of files.
The easiest way to do create the bbs ads database involves collecting
all those nasty ads together and zipping them up into one zipfile.
You can use whatever name you like for the ZIP. Then you run the
program ZDCSBA from the directory containing all the ZDCS files on
that zipfile.
This is easy to see with an example. If your collection of files to
be added is in a zip called BBS-ADS.ZIP, then the command you issue to
create the bbs ads database with these files is:
ZDCSBA BBS-ADS.ZIP
You don't have to do anything else. The program handles everything
for creation of the bbs ads database.
If you want to create a new bbs ads database in the future, just
delete the old database file (ZDCS-BBA.NDX) and follow the steps for
creating a new bbs ads database. If you don't delete the old
database, then the new ads will be added to the old ones in the
database, which is an easy way to add new bbs ads.
Updating the BBS Ads Database
-----------------------------
There are three easy ways to add new bbs ads as they are foisted off
on the bbs community.
1. You can run ZDCSBA on a new zipped collection of ads when you
already have an existing bbs ads database. This is good even if
the zipped collection only has a single file in it. ZDCSBA works
just the same way here as it does for creating a new database,
with one exception: the files being processed are added to the
existing database. That's all there is to it.
Example. Let's say that you've collected some bbs ads from your
board, your own downloads, and even from friends. You zip
- 14 -
the entire set together and call it JUNK.ZIP. To include
these files in the bbs ads database, just issue the command:
ZDCSBA JUNK.ZIP
2. You can run ZDCSBA on an individual bbs ad without bothering to
zip it up first. This is a quick way to catch a new ad as soon
as you are lucky enough to get it.
Example. You've just come across an obnoxious get-rich-quick scheme
called MOOLAH.TXT. To put this file in the bbs ads
database, just issue the command:
ZDCSBA MOOLAH.TXT
3. You can even run ZDCSBA to include a file in the bbs ads database
when you don't have the original file, as long as you have the
CRC32 value for it from PKZIP. Just run ZDCSBA on the CRC32
value immediately preceded by a dollar sign $ instead of on the
file itself.
Example. You don't have the original ad anymore, but you have a
record that the CRC32 of a new bbs ad is E1E10999. To put
this file in the bbs ads database, just issue the command:
ZDCSBA $E1E10999
You can use any or all of these methods to add new records to the bbs
ads database. If you have an assortment of zipped collections, a
couple of unzipped ads, and a few ads known only by CRC32, the data
can all be entered into the same database without having to use the
same technique for each individual file.
Selecting Deletion or Flagging of BBS Ads
-----------------------------------------
If you have built the bbs ads database, ZDCS will automatically detect
its presence and will compare uploads against the records in the bbs
ads database. What ZDCS does with any matches it finds depends on how
you set up the configuration file.
In line 6 of the ZDCS configuration file ZDCS.CFG, you typed one of
two letters: Y or N. The letter Y tells ZDCS to delete the offending
bbs ad. The letter N tells ZDCS not to delete it, but to just flag it
as a bbs ad.
When a bbs ad in an upload is merely flagged, that information is
still displayed to the caller and recorded in the log for the upload
file checker. If you wish, you can go back and manually delete such
files later on.
You can change your mind about using deletion or flagging to deal with
bbs ads at any time. All you have to do is edit that one line in the
configuration file. No other changes need to be made to run with the
new option.
- 15 -
ALLOWED DUPLICATES
------------------
Allowed duplicates are those files that may already be present in your
file collection, but which you do not want ZDCS to treat as
duplicates. You can establish a list of files that are allowed
duplicates. When ZDCS evaluates an upload, these allowed duplicates
will not be flagged as dupes. They are protected from deletion, even
if you have configured ZDCS to delete duplicate files. Furthermore,
when ZDCS evaluates the actual percentage of duplicates in an upload,
these files are not counted.
The reason behind the concept of allowed duplicates is that there are
some files that reappear frequently as critical components of more
than one major shareware package or upgrade version. These might be
order forms, registration guides, general text files, small utilities
or other files. Two common examples are the OMBUDSMN.ASP file found
in all ASP-ware and the VALIDATE.COM program included in every
revision of McAfee's SCAN package. Both files turn up unchanged in
many zipfiles. It would be misleading to treat these files as
ordinary duplicates.
There are two functional advantages to using allowed duplicates.
First, by not counting allowed duplicates in the actual percentage of
dupes calculation for new uploads, your system does not "penalize" a
zipfile for having one or more of these files when it comes to
deciding whether to accept or decline the file. This is especially
significant if you have set a pretty high standard for the
"uniqueness" of uploads that your bbs will accept.
The second advantage is for those systems that configure ZDCS to
delete duplicate files. The use of allowed duplicates can protect
popular shareware packages from being stripped of important files.
Not only does the deletion of such files alter the content of the
author's shareware package, but it also destroys any AV stamp the
author placed on the work.
Whether or not this option is enabled is controlled by the presence or
absence of an ASCII text file called ZDCS.ADN in the ZDCS directory on
your system. This file is one that you create with any text editor to
list all the files, one per line, that are allowed duplicates. You
can designate an allowed duplicate by either its file name or its
CRC32.
To specify an allowed duplicate by name, just type the dollar sign $
followed immediately by the name of the file with its extension. This
preserves a file with a distinctive name (like OMBUDSMN.ASP) even if
it undergoes some revisions.
To specify an allowed file by its CRC32 value, type the pound sign #
followed by the CRC32 for the file. If you don't know the CRC32
offhand, you can get this information from PKZIP. Issuing the PKZIP -
V command on any zipfile gives you the CRC32 values for each
individual file inside the zipfile.
The ZDCS.ADN file may be up to 256 lines long, but must contain no
blank lines and no blank spaces. You can edit, delete or recreate
- 16 -
this file at any time without having to set any other switches, alter
the configuration file, or make any other changes to the setup of
ZDCS.
While the expected application of the allowed duplicate feature is
with zipfiles, it does work for GIF files too. If a GIF is included
among the allowed duplicates, then a repeat upload of the same GIF
will not be flagged as a duplicate, and the upload will be accepted.
THE UPLOAD FILE CHECKER
-----------------------
Purpose
-------
Once you have created the initial ZDCS duplicate file database, you
can get the bbs to check all uploaded zipfiles and GIFs against it
from then on. This is done by processing the uploaded zipfiles and
GIFs with the upload file checker ZDCSFC as the files are received.
The primary purpose is to catch uploaded files that duplicate ones
already on the bbs.
ZDCSFC can also flag bbs ads, as described in that section of this
manual. Depending on whether you have enabled any deletion options in
the configuration file, ZDCSFC can also perform deletions of duplicate
files and/or bbs ads. With the use of a list of allowed duplicates,
also described in a separate section in this manual, the file checking
performed by ZDCSFC can be made even more sophisticated.
Function
--------
ZDCSFC works with the CRC32 value for each file in an upload. If the
uploaded file is a GIF, it first calculates the CRC32 for that file.
If the uploaded file is a zipfile, ZDCSFC reads the value of the CRC32
for each individual file in the zipfile.
Then ZDCSFC compares the CRC32 for each file in the upload against the
ZDCS database. If you have created the bbs ads database and left it
in the ZDCS directory on your system, ZDCSFC also compares the files
in the upload against the bbs ads database.
The results are displayed file by file for the caller after the upload
has been received and processed, providing real time feedback. The
same information is also available to the sysop as part of the ZDCSFC
log file. The complete path and file name for this log were specified
as line 5 of the configuration file.
ZDCSFC also calculates the actual percentage of duplicates in any
upload and compares this against the maximum percentage you have set
for the board. On the basis of this comparison, an upload may be
accepted or declined. Declined uploads are not deleted from the
system. PCBoard moves a declined upload to your private directory for
sysop review.
- 17 -
The calculation of percent duplicates is covered in more detail in a
separate section of this manual dealing with maximum and actual
percent duplicates in an upload.
Maximum and Actual Percent Dupes in an Upload
---------------------------------------------
ZDCSFC calculates the actual percentage of duplicate files in the
upload. Since a GIF is a single file, it will either be 0% (not a
dupe) or 100% (a dupe). For zipfiles, this actual percentage can vary
anywhere between 0 and 100. ZDCSFC compares this actual percentage
against the maximum percentage you set in line 4 of the configuration
file.
If the actual percentage is lower than the maximum, the upload is
accepted. If the actual percentage is equal to or higher than the
maximum you specified, the upload is declined. PCBoard moves these
declined files into your private upload directory, where you can
review them. Some callers have become frantic at the idea that a
"declined" file is thrown out. This does *not* happen with ZDCS.
If you want to bypass this filter, set the percentage to 100. This
permits a duplicated GIF or a zipfile with nothing but duplicates to
pass the filter and never be declined. At the other extreme, you can
set the percentage to 0. This effectively requires that the uploaded
GIFs and zipfiles have no duplicates at all, or they will be declined.
The section in this manual that describes the configuration file
specifies that the maximum percentage of dupes in line 4 must be a
whole number, not a decimal. This is because ZDCS deliberately
truncates any decimal from this value. ZDCS does not round up or down
but performs a straightforward truncation on both the actual and the
maximum percentage of dupes. This was done for reasons of memory
management. An example may make it clearer how this approach works in
actual practice.
Suppose that the maximum percentage of duplicates is set to 20. An
upload with an actual percentage of 19.9 would be truncated down to
19. ZDCS would see that 19 is less than 20 and would accept this
upload. If the actual percentage were rounded instead of truncated,
it would be 20 instead of 19. ZDCS would then decline this upload
because the actual percentage would not be less than the maximum
percentage of dupes.
Updating ZDCS Database(s) After an Upload
-----------------------------------------
ZDCSFC automatically updates the duplicate files database with the
CRC32s of all uploads. You do not have to do anything special to
include this information in the database for comparison with future
uploads.
ZDCSFC does not modify the bbs ads database at all. It's still not
smart enough to recognize a bbs ad until you've pointed it out first -
but it does remember them next time it sees them. Please see the
section in this manual on the bbs ads database for more information.
- 18 -
Calling ZDCS from EXZTEST
-------------------------
ZDCS should be used in conjunction with a file integrity checker. If
you have decided to use EXZTEST, then how you call ZDCSFC to process
new uploads depends on which version of EXZTEST you are running.
The newer versions of EXZTEST provide a seamless integration with ZDCS
that takes care of calling the ZDCSFC module, feeding it the
information it needs, and completing any file deletions that you have
enabled. This is available in EXZTEST version 2.0 and higher.
A sample PCBTEST.BAT file that calls the newer (2.x) version of
EXZTEST is included with this package under the file name PCBTEST.ALT.
For more information on the ZDCS integration implemented in EXZTEST
2.x, please see the documentation for EXZTEST.
The older versions of EXZTEST before version 2.0 can still be used in
conjunction with ZDCSFC to check new uploads. Instead of relying on
EXZTEST to call ZDCS, you must call the upload file checker from the
PCBTEST.BAT file directly.
A sample PCBTEST.BAT file that calls the older (1.x) version of
EXZTEST and then calls ZDCSFC is included with this package under the
file name PCBTEST.BAT. You'll have to modify it a bit to reflect the
pathnames and other details specific to your system, but it's a good
general guideline.
More detailed information can be found in this manual in the section
on calling ZDCSFC from the PCBTEST.BAT file. This section is more
general and addresses the use of any file integrity checker.
Calling ZDCSFC from the PCBTEST.BAT File
----------------------------------------
ZDCS should always be used in conjunction with a file integrity
checker. When processing new uploads, the file integrity checker
should be called from the PCBTEST.BAT file before ZDCSFC. If you are
using the PCBDescribe feature in PCBoard 14.5a, that command should be
used first of all.
To use ZDCS with the file integrity checker of your choice (even if
it's just PKZIP -T), there are six basic pieces you need to include in
the PCBTEST.BAT file.
1. Include the following three lines to clean out old copies of
these files left over from processing other uploads:
@IF EXIST PCBFAIL.TXT DEL PCBFAIL.TXT
@IF EXIST PCBPASS.TXT DEL PCBPASS.TXT
@IF EXIST ZDCS-DEL.LST DEL ZDCS-DEL.LST
PCBFAIL.TXT contains information displayed to the caller when an
upload is declined. PCBPASS.TXT contains the information
displayed when the upload is accepted.
- 19 -
ZDCS-DEL.LST is a list of files targeted by ZDCS for deletion
from an upload. These might be duplicate files, bbs ads, or
both. If you haven't enabled any deletion options in the
configuration file, or if there were no files meeting the
criteria for deletion, ZDCS would not generate the ZDCS-DEL.LST
file for that upload. By deleting the files here, you make sure
that no old deletion instructions are hanging around when a new
file is processed.
2. Call your file integrity checker to process the upload. The
exact command to do this will depend on which integrity checker
you are using. The most basic and simple one of all is to make
use of this ability in PKZIP by using the -T switch, but there
are other programs to do file integrity checking and more.
3. Call ZDCSFC to check the upload. The appropriate command is:
ZDCSFC %1 %2 %3
If you are running a version of PCBoard older than the May 22,
1991 version 14.5a, PCBoard will provide only two parameters.
Since the third one is absolutely necessary for pre-testing, that
means you must disable the pre-testing feature. There is a way
to get around this limitation if you absolutely must. It's not
exactly elegant, but it does work. If you need to know more
about it, contact us in one of the conferences mentioned in the
support section of this manual.
If you are using a newer version of PCBoard but you don't want to
use the pre-testing right now, it is still strongly recommended
that you leave the third parameter in place. It does no harm and
could save you some grief if you change your mind in the future.
At the end of processing by ZDCSFC, you might have a new file
called ZDCS-DEL.LST. This depends on whether you have enabled
any deletions and on whether there were any files in the upload
that met your criteria for deletion. If no files are to be
deleted, this control file won't be created.
4. Skip to the end if there are no files to be deleted:
IF NOT EXIST ZDCS-DEL.LST GOTO END
This line tests for the presence of the ZDCS-DEL.LST file. If
the file is not found, it means that there are no deletions to be
performed on the current upload.
5. Perform the deletion of files specified by ZDCSFC:
PKZIP -D %1 @ZDCS-DEL.LST
Note that the actual deletion is done by PKZIP. ZDCSFC creates
the control file ZDCS-DEL.LST to specify the deletions and
PCBTEST.BAT passes this file to PKZIP at this point. This is
also the only time an existing AV stamp on an upload is affected
by ZDCS. More information on AV stamp integrity is available in
a separate section of this manual.
- 20 -
6. Mark the end:
:END
When there are no files to be deleted, this statement marks the
location where control in the PCBTEST.BAT file is passed.
DOS Error Levels
----------------
In addition to creating the required PCBPASS.TXT and PCBFAIL.TXT
files, ZDCSFC also sets the DOS error level when it exits. These
levels are:
0 No duplicate files were found within the upload.
1 Some duplicates were found, but the upload passed the percentage
test.
2 Too many duplicates were found, and the upload failed the
percentage test.
3 Every file within the upload was a duplicate.
4 There is no number 4. (Reserved for future expansion.)
5 Upload checking by ZDCSFC was aborted. Please see log for error
message.
PROCESSING LOCAL UPLOADS
------------------------
There probably isn't a sysop alive who hasn't scavenged new files for
the board and uploaded them locally. ZDCS offers two ways of
processing these local uploads. The first method is to use the
database builder ZDCSDB; the second method is to use the upload file
checker ZDCSFC that processes all the regular uploads from callers.
In the first case, ZDCSDB can be used to add all the new files to the
database. Just like the initial creation of the ZDCS database or
subsequent database merges, ZDCSDB adds the CRC32's of the new files
to the database. The duplicate report generator ZDCSDR can still be
used to tell you about duplicates in the database. A disadvantage to
this method is that it does not give the sysop the same kind of
immediate feedback for a local upload as the sysop and caller get for
a regular upload. It also does not delete bbs ads or duplicate files
(if you have those options enabled).
To use ZDCSDB to process local uploads, you only have to change the
first line of the ZDCS configuration file. This line points to the
paths of the files to be processed when running the database builder.
(The line contains the full path / name of the text file that lists
all the directories whose files are to be processed.) Change this to
- 21 -
point to the uploads directory, a holding directory, or whatever
directory will hold the local uploads to be processed. Now just run
ZDCSDB and the files will be added to the existing database.
This method does not interfere with the realtime upload checking
performed by ZDCSFC on regular uploads, because the path information
in line 1 of the configuration file is not used by ZDCSFC. If you've
decided on a permanent holding directory, you won't even have to touch
the configuration file again to process any future uploads. Just
remember to run ZDCSDB before posting new files on the board.
This is a good time to repeat one note of caution: ZDCS does not do
any integrity testing. ZDCS is intended to work with your choice of
integrity tester, not to replace it. If you have not used an
integrity checker on the local upload files (why not?), then you need
to use the T (for Test) switch. Instead of using the command ZDCSDB
to run the database build, use the ZDCSDB T command.
This calls PKZIP to do an integrity check on the zipfiles. Any file
that fails this test is not processed by the database builder. There
will be a message in the database building log file ZDCS-DBB.LOG for
each such damaged file.
Don't use the T switch without reason - it adds another step and that
slows the processing down tremendously. But if you have a collection
of files that haven't passed through an integrity checker, like
prospective local uploads, ZDCSDB T is the way to deal with them. If
ZDCSDB has to skip a file because it's flagged as damaged, that
information will show up in ZDCS-DBB.LOG.
The second method mimics regular uploading to get the local uploads
processed by the upload file checker ZDCSFC. This is handled via a
small utility and batch file combination available free on The Hacker
Central BBS as LOCALUP.ZIP. You'll have to modify the batch file for
your own system, of course.
The result is that the local uploads will see the same processing that
a regular caller's upload does on your board - including your usual
file integrity checker, bbs ad deletion (if enabled), duplicate file
deletion (if enabled), and all the rest. This has the additional
advantage of giving you the ZDCSFC log file entries for the local
uploads, which gives the status of the files inside the zipfiles:
duplicates, bbs ads, allowed duplicates, etc.
Local upload processing is an area targeted for future enhancement in
ZDCS.
PRE-TESTING
-----------
Callers can pre-test an upload to find out how it compares to files
already on the bbs in terms of duplicates and bbs ads. The idea is to
see how the prospective upload compares to existing files in terms of
duplicates, bbs ads, etc. before actually uploading the full file to
the bbs. It can save both caller and sysop a certain amount of
frustration.
- 22 -
The procedure is wonderfully simple from a caller's point of view and
requires nothing that might tax a relatively novice uploader's skills.
There are no special files to download, no complicated operations to
get right, no arcane rituals to perform. This makes it more likely
that callers will take advantage of the pre-test feature.
Included in this release of ZDCS is a sample bulletin that can be
posted by the sysop on the bbs to explain to callers the ZDCS pre-
testing feature for zipfiles. The bulletin holds the caller's hands,
figuratively speaking, through the whole process. If your bbs permits
the uploading of SFX or GIF files, you might want to add those
initials where you see ZIP in the bulletin. A ZIP-only bbs can use
the canned bulletin right from the package.
To pre-test an upload called GOODSTUF.ZIP, the caller simply issues
the DOS command:
PKZIP -V GOODSTUF.ZIP > ZDCSTEST.CHK
This command is not case sensitive, so it's pretty hard to make a
mistake with it.
The caller then uploads the file ZDCSTEST.CHK directly to the bbs.
ZDCS will read the information in it and will give the caller a
breakdown of the zipfile's contents, noting which individual files are
duplicates or bbs ads. This output looks just like the response a
caller gets after actually uploading the zipfile, with one exception:
instead of telling the caller whether the file is accepted or
declined, ZDCS tells the caller what percentage of files in the upload
are duplicates, and also gives the maximum percent duplicates which
the sysop has allowed for the board. Now it's up to the caller to
decide whether to upload or not.
Enabling the pre-test capability is not tough on the sysop, either.
The key to whether or not pre-testing is permitted on your bbs lies in
the UPSEC file. If you want to prevent callers from using pre-
testing, just disallow uploads that have the .CHK extension. On the
other hand, if you want callers to have the option to pre-test, make
sure that .CHK files may be uploaded to
the system.
There's one more thing to check before letting your callers use the
pre-testing feature. If you are not using EXZTEST version 2.0 or
higher, remember to make sure that the PCBTEST.BAT file contains the
following line:
ZDCSFC %1 %2 %3
This is already discussed elsewhere in this manual as part of the ZDCS
installation. That third parameter is necessary to let the pre-
testing work for more than just the first caller. If you leave it
out, the first ZDCSTEST.CHK file uploaded to your system will work
just fine, but all subsequent attempts to pre-test will be told that
the ZDCSTEST.CHK file is a duplicate! And that means they won't be
recognized as attempts at pre-testing.
If you are using EXZTEST version 2.0 and up as your file integrity
checker, you won't have a ZDCS line in the PCBTEST.BAT file. That's
- 23 -
because EXZTEST has a seamless interface to ZDCS and calls it
directly.
Since this third parameter is critical to the proper functioning of
the pre-test option, you must have a recent enough version of PCBoard
that supports the third parameter. The earliest version of the board
code that satisfies this requirement is PCBoard 14.5a from May 22,
1991. This version and anything after it can handle ZDCS pre-testing.
If you track your callers' uploads and downloads, you might like to
know that the ZDCSTEST.CHK will not count as an upload. According to
ZDCS and PCBoard, the ZDCSTEST.CHK file always "fails" validation and
is not credited to the caller as an upload. The bulletin does mention
that pre-testing doesn't count as an upload, just in case some of your
callers are too enthusiastic about getting upload credits.
ZIPs WITHIN ZIPs
----------------
Processing of zipfiles contained within zipfiles is accomplished with
some caveats. Zipfiles within a zipfile are only checked one level
deep. The simplest explanation is to look at an example.
ZIP A contains assorted files and ZIP B. In turn, ZIP B contains more
files and another ZIP, C. ZIP C contains still more files. How does
the whole melange get processed?
All the files in ZIP A and in ZIP B have their CRC32 signatures
entered into the duplicate files database. If you have configured
ZDCS to delete any files (dupes or bbs ads), then those deletions are
done automatically only for the individual (non-zipped) files inside
ZIP A. No files are deleted from inside ZIP B. Of course, all the
duplicates in ZIP B are still listed in the log, so you do know about
them and you can decide whether to remove them manually from your file
system.
What about ZIP C? That's easy: any zipfile embedded more than one
level deep in the uploaded zipfile (and C is two levels deep) is not
processed as a zipfile at all. No file deletions, no CRCs, nothing.
In fact, *any* file that has the .ZIP extension will not be treated as
an individual file in the database or the log file. The reason is
simple: it's the individual files inside the zipfiles and not the
zipfile itself that are important when looking at duplicates.
(Besides, it plays havoc with the percentages.)
CD-ROMs
-------
CD-ROMs have been growing in popularity among bbs's because they
enable a large, static collection of files (such as shareware
programs) to be available for downloading. ZDCS works just fine with
CD-ROM drives, but the sheer size of these disks means that it can
take quite a while to build the ZDCS database for a CD-ROM.
- 24 -
To make life easier, a new ZDCS database merge utility is now
available on The Hacker Central. By merging databases, it is possible
to take advantage of the existence of prebuilt ZDCS databases for the
more popular CD-ROMS. This lets you merge a prebuilt database into
your own bbs's database in a fraction of the time it would take to add
all the CD-ROM information from scratch.
Prebuilt ZDCS databases for some CD-ROMs are available free on The
Hacker Central BBS. If you have a CD-ROM that doesn't already have a
prebuilt ZDCS database, please consider creating a ZDCS database for
the individual CD-ROM and sharing it with other ZDCS users by
uploading it to The Hacker Central. During 1992, we are offering a
free ZDCS registration as a thank you to anyone who uploads such a
database.
AV STAMP INTEGRITY
------------------
Registered versions of PKZIP can "brand" a zipfile with the unique AV
(authenticity verification) number of the registered owner. The AV
stamp is used by some shareware authors for their official
distribution packages. It is also abused by some people, including a
few bbs sysops, sad to say. Whether ZDCS will always leave that AV
stamp intact is determined by the configuration of ZDCS.
The first possibility is that ZDCS is configured never to delete any
files, whether duplicate files or bbs ads. In this case, the original
AV stamp is always retained.
The second possibility is that ZDCS is configured to delete files, but
no files need to be deleted from a particular zipfile. In that case,
any AV on the original zipfile is left intact.
The third possibility is that ZDCS deletes one or more bbs ads.
Whether the AV stamp is left intact depends on whether the bbs ad was
included in the zipfile inside or outside the AV branding. If the
offending bbs just added their advertising to the outside of an AV
stamped zipfile and left the original author's AV stamp in place, ZDCS
will do the same. If that bbs added their overhead file and then
rezipped the entire package with their own AV stamp (don't laugh,
there's at least one large bbs that's done this), then they have
destroyed any original AV stamp. ZDCS does not preserve the AV stamp
in this case.
The fourth possibility is that ZDCS deletes one or more duplicate
files. If any deleted file was included in the AV branding, then the
AV stamp is not preserved. If all deleted files were outside the
original AV stamp, then the AV stamp is intact. (This could happen
with some advertising file that you've already been hit with but which
doesn't appear in your bbs ads database.)
This last possibility deserves a little thought from the shareware
author's point of view. It is not uncommon for a new version to
contain some files that are unchanged from the previous version. If
- 25 -
these files haven't been flagged as allowed duplicates, it is possible
to choose to configure ZDCS so that these files are removed. If that
happens, the AV stamp is also removed. This is in keeping with the
ideal that the stamp should represent an authenticity verification
from the original author.
ZDCS is quite capable of handling and retaining all AV stamps. This
is a matter of choice on the part of the sysop and is the direct
result of the simple configuration options.
ACCURACY OF THE CRC32 METHOD
----------------------------
In order to detect duplicates, some unique signature is needed for
each file. While the only truly unique signature is as long and
cumbersome as the file itself, there are more manageable alternatives.
Any algorithm is a trade-off between these two factors: manageability
and accuracy. The more unique signatures a method can generate, the
smaller the possibility that two different files might generate the
same signature.
ZDCS makes use of the CRC32 used by PKZIP. This signature has a total
of eight "places", each filled by independent assortment with one of
sixteen different values. That makes for a total of
16*16*16*16*16*16*16*16 (16 to the 8th power) unique signatures. With
over four and a quarter billion possible signatures, this method has
good accuracy.
It is possible to use other methods to permit an even greater number
of unique signatures. The longer signatures and different calculation
methods carry a corresponding cost in complexity, disk space and
processing speed.
MEMORY LIMITS
-------------
The ZDCS upload file checker ZDCSFC.EXE requires a minimum of 384K of
free memory (RAM).
There is an internal limit of 650 files within a zipfile for
processing by both the upload file checker ZDCSFC.EXE and the database
creator ZDCSDB.EXE. If there are more files inside the zipfile, ZDCS
will handle the first 650 it encounters and will leave a message in
the log. There is nothing to reset and no effect on processing the
next file ZDCS encounters. These limits are imposed by the DOS 640K
Limit.
TROUBLE-SHOOTING GUIDE
----------------------
There is no better help for solving problems than experience,
especially someone else's experience with the same problems.
- 26 -
ZDCSDB (the database build) is crashing.
---------------------------------------
There is a log file ZDCS-BBA.LOG that collects messages when ZDCSDB is
used to create a new duplicate file database or add to an existing
one. Read the messages in the log file to find out why the database
build program is crashing.
Most of the time, the culprit is a damaged zipfile. ZDCS normally
depends on the fact that most sysops use a separate integrity file
checker on their boards. If a corrupted zipfile gets into the
collection of files being processed into the database, it can bring
things to a sudden halt.
The recommended solution is to process all files with a file integrity
checker first. However, there is a second alternative using just
ZDCS. You can use the T (for Test) switch to tell ZDCSDB to call
PKZIP and perform a file check first. Only those files passed by this
check are processed by the database builder. The syntax is simple:
ZDCSDB T instead of ZDCSDB. The penalty in speed for doing it this
way is very significant, which is why it is not the usually preferred
solution.
ZDCS is reporting a device I/O error (in any module).
----------------------------------------------------
So far, we have seen this happen twice. Each time, ZDCS was running
on a LAN. In both cases, the problem was traced back to bad spots on
the hard drives.
The best medicine for this error message is to run Norton Disk Doctor,
Spinrite, or something similar on your hard drive system.
ZDCS reports memory corrupt errors (in any module).
--------------------------------------------------
This sounds like an earlier version of ZDCS, probably version 1.62 or
earlier. Those versions required more free RAM and sometimes ran out
of memory, which resulted in the misleading "memory corrupt" messages.
Time to stop reading the docs and get this newer release (version
1.65) installed! The improved memory management in ZDCS 1.65 should
solve your problems.
ZDCS reports path/file access errors.
------------------------------------
One possibility is that you are using an earlier version of ZDCS on a
LAN *and* you are not loading SHARE before running ZDCS. This is most
likely to happen on a Novell Netware LAN. You can solve this one by
loading SHARE before you run any of the ZDCS modules on a network.
Better still, you can upgrade to this version 1.65 of ZDCS.
If you have loaded SHARE high under DOS 5.0, you might still be
experiencing the same symptoms. There seems to be a problem with
- 27 -
loading SHARE high; it's not anything specifically to do with ZDCS.
Try loading SHARE low instead.
Another possibility is that you haven't specified enough files in your
CONFIG.SYS file. Because ZDCS makes such heavy use of disk I/O, you
should make sure that at least 30 files are specified in CONFIG.SYS
(FILES=30).
Of course, if you are trying to use the upload file checker ZDCSFC
before you've built the duplicate file database with ZDCSDB, you
shouldn't be surprised if ZDCS acts a little weird. Try building the
database first and then comparing new files against it. Things work
much better that way.
QEMM exception 13 errors occur when running ZDCS.
------------------------------------------------
There are five possible reasons for this error message.
1. You're running an earlier version of ZDCS, probably version 1.62
or even earlier. Those versions required more free RAM and
sometimes ran out of memory, which could result in the QEMM error
messages. Upgrade to the current version 1.65 of ZDCS, which has
improved memory management.
2. Your QEMM command line is incorrect. The problem isn't with ZDCS
itself, but ZDCS is likely to be the first program to notice this
QEMM error because of the amount of data ZDCS needs to juggle.
The result is that ZDCS is able to write its data someplace it
shouldn't. To solve this problem, check that all areas of memory
that should be excluded from QEMM are being properly excluded. Be
especially careful if your system has super VGA, SCSI devices or a
LAN: all of these devices usually require specific areas of
memory to be excluded from remapping by QEMM.
3. You have one or more "mis-behaved" device drivers. You can check
out this possibility by removing all drivers from your CONFIG.SYS
file, rebooting the system, and trying ZDCS again. One example of
this type of problem was with an older version of the Hyperdisk
cache.
4. You believed the QEMM optimize program when it told you to get rid
of DOS stack space. Put it back! QEMM is egocentric enough to
believe it can handle all types of stack requirement for any
application properly. Let's just say that putting back the DOS
stack space solves this problem nicely.
5. You are trying to use the "stealth technology" feature of QEMM
6.x. There have been persistent reports of problems traced to
interactions between stealth and various applications from other
vendors. It wouldn't be at all surprising if ZDCS experienced the
same trouble with stealth. This is not a bug in ZDCS. Contact
Quarterdeck tech support. In the meantime, disable stealth by
removing the ST:F or ST:M from the DEVICE=QEMM.SYS line in the
CONFIG.SYS file.
- 28 -
REGISTRATION
------------
ZDCS is a fully functional shareware package. There are no critical
limits, crippled features or "drop dead" dates. The only difference
in the unregistered version is a brief message to sysop and caller
that the version is unregistered.
A customized ZDCS.KEY file is available to registered users to change
the "unregistered" message to show registration in the name of <BBS
Name> or to eliminate the display of any message at all. Registration
entitles the user to a license for use of the Zipfile Duplicate
Checking System on one bbs, no matter how many nodes it has. This
includes all future versions of ZDCS. No additional fees will be
charged for registration of future versions of this product.
If you try ZDCS out on your system and decide that you want to
continue using it, please register your copy by sending a check for
$25.00 (US). Include the name of your bbs (up to 25 characters) as
you want it to appear to your callers. An order form (ORDER.FRM) is
included in this release for your convenience. We will prepare a key
file for your copy of ZDCS with this information in it to replace the
"unregistered version" line in the display. This keyfile will be
available for download from The Hacker Central BBS.
If your bbs is located in the continental United States or Canada, you
may take advantage of the keyfile delivery service any time during
1992. For an additional $5.00 (US), we will prepare your keyfile and
upload it to your bbs. Just set up an account in the name of ZDCS
SUPPORT with the password KEYFILE and with sufficient security to
upload a private file for the sysop. When you send in your
registration of $25.00 plus $5.00, remember to include the name and
number of the bbs.
As of 1991, ZDCS can now be registered online at The Hacker Central
BBS with a valid Visa or Mastercard. Look for the ZDCS script
questionnaire in order to register. The keyfile delivery service is
also available as part of the online registration. Processing usually
takes from one to three days.
If you prefer to register by check or money order, please make your
registration check payable to Michael W. Cocke and mail it along with
the necessary information (like a completed order form) to:
Michael W. Cocke
11 Cedar Road
Montville NJ 07045-9582
Please be assured that ZDCS will continue to receive support through
future revisions. For over a year now it has been in use on its home
board, The Hacker Central BBS. When the programmer is also a sysop
and has to live with the results of that programming work every day,
you can be sure that the support will be there!
- 29 -
SUPPORT
-------
ZDCS product support will be handled on the ZDCS home board and via
echoed conferences. Questions, comments and discussion of ZDCS are
welcome in the following conferences:
SHAREWARE (ILink)
ZDCS (Intelec)
ZDCS (Hacker Central, also echoed by many ILink bbs's)
Additional product support in the form of downloadable files
(including registration keys) is available on the ZDCS home board, The
Hacker Central BBS.
First time access to The Hacker Central BBS must be done via the
public node. After completing the new user and visiting sysop
scripts, all sysops and ZDCS users (whether or not they have
registered ZDCS yet) are granted access to the two private nodes,
including the high speed line. The Hacker Central phone numbers are:
Node 1 201-334-2555 Public 2400
Node 2 201-316-8840 Private 2400 MNP
Node 3 201-335-9343 Private USR HST/DS
Please leave a comment to sysop on your first call indicating your
interest in ZDCS in order to get access to the ZDCS support
conference. Requests are usually handled in a day or two.
Another aspect of ZDCS product support is the collection of utilities
and pre-built CD-ROM databases available for download from The Hacker
Central BBS for all ZDCS users. This collection is constantly growing
in response to suggestions, requests and cries for help. A few of the
special utilities are described below.
The database merge module enables you to combine more than one ZDCS
duplicate file database into a single database. This has the
tremendous advantage of permitting pre-built ZDCS databases to be
created for the more popular CD-ROMS. Once accomplished, a single CD-
ROM database can be merged into any number of bbs databases in a
fraction of the time it would take to create the CD-ROM database from
scratch.
Pre-built CD-ROM databases are available on The Hacker Central. We
encourage anyone who has a CD-ROM without a prebuilt database to
create the ZDCS database for the CD-ROM alone and to upload it to The
Hacker Central to share with other sysops. To encourage this, we are
offering a free ZDCS registration to anyone who uploads a new CD-ROM
ZDCS database to The Hacker Central BBS during 1992.
The FWK to ZDCS converter will take the data from an FWKCS(tm)
database and convert it to a ZDCS database without requiring the
original source files to be available. If you have tried FWKCS(tm)
and want to evaluate or switch over to ZDCS, this makes the process
easier and faster.
If you have used an earlier version of ZDCS (1.5x) and haven't
upgraded to the smaller and faster version 1.6, there is a free
- 30 -
upgrade kit available on The Hacker Central to help you do that
without starting over again from the beginning.
The database editor permits you to modify or delete records in a ZDCS
database. It is intended primarily as a debugging tool. It is
strongly recommended that a backup copy of the database be made before
embarking on any editing.
A small local uploads package has been put together to make procesing
local uploads easier for the sysop who also scavenges files for the
board. Additional work is also in progess for improving the handling
of local uploads.
Development on other utilities continues. One module in progress is
geared toward ferreting out instances of bbs ads that might have crept
into the database. With some bbs's changing their ads frequently,
this can happen before the sysop realizes it and flags the offending
ads. Suggestions for other new utilities are always welcome.
All ZDCS utilities to date are free but are only available on The
Hacker Central BBS. Please do not upload these modules to other bbs's
or distribute them as part of the shareware ZDCS package.
A final note about ZDCS support concerns documentation. We feel that
clear and helpful documentation is an important part of a sysop
utility package like ZDCS. In addition to the plain text files
included in this package, we also have the documentation available in
WinWord 2.0 with full use of styles, headers, outline processing and
other features. If you are interested in obtaining a more
presentation-quality copy of the documentation or in discussing
conversion to other formats such as WordPerfect, please contact us on
The Hacker Central BBS. If you have any questions, comments or
suggestions concerning the documentation, we would like to hear from
you, too.
FUTURE ENHANCEMENTS
-------------------
ZDCS is a dynamic product. In response to suggestions, comments and
the changing bbs scene, ZDCS continues to be upgraded. Future
enhancements under consideration or in progress include support for
other archive formats besides PKZIP, improved local upload handling,
and database path updating.
Path updating would keep the file location information in the ZDCS
database current even after files are moved around in the file system.
Right now, the essential identification information (the CRC32
signature) remains fine, but the path pointing to the file becomes out
of date when the file is moved. This is usually the case with uploads
that are moved from a recent uploads directory to another destination.
Path updating would allow the sysop to locate these and any other
moved files more quickly by keeping the path information current as
well.
- 31 -
ACKNOWLEDGEMENTS
----------------
Thanks to all the beta testers who have helped to make ZDCS a better
and smoother program. Their patience and input are definitely
reflected in the final product.
Thanks to Tom Hughes for his prowess as a bug hunter extraordinaire,
and for wanting a database merge utility. That module was a great
idea, Tom!
Thanks to David Terry for several clever ideas on better ways to do
things. That kind of help is always appreciated.
Thanks to Bob Jacobson for his thoroughness in finding a particularly
persistent and recalcitrant bug during the recent long beta cycle. It
(the bug, not Bob) has been sighted and squashed.
Special thanks to Andy Keeves for pointing out a better hammer in the
toolbox. It was by making use of his suggestion for a different
sorting routine that the duplicate report generator was made
significantly faster.
Finally, thanks to everyone who had the patience to wait while version
1.65 went through six months of intense beta bashing.
COPYRIGHTS AND LEGAL STUFF
--------------------------
ZDCS (Zipfile Duplicate Checking System) is copyright (C) 1991 by
Michael W. Cocke. ZDCS 1.65 is fully functional shareware. It may be
freely copied and distributed, provided that no files in this package
are removed or altered in any way.
The individual documentation files ZDCS-REF.TXT, ZDCS.BLT and
ZDCSWALK.TXT are copyright (C) 1991 by Evelyne Stalzer. Permission is
granted to distribute these files as part of the complete ZDCS
shareware package. Permission is also granted to bbs sysops to modify
ZDCS.BLT for use as needed, including as a bulletin, message or
downloadable help file.
ZDCS.KEY is the individual registration key file and may not be
copied, distributed or otherwise shared with individuals beyond the
registered user.
Neither Michael W. Cocke, Evelyne Stalzer nor MWC Enterprises will
accept responsibility for the function, failure to function, or side
effects of any function of the Zipfile Duplicate Checking System
(ZDCS). ZDCS is provided in good faith, but its use is solely at the
risk of the operator.
EXZTEST and EXZIP are copyright Andy Keeves.
FWKCS is trademark and copyright Fred Kantor.
Lantastic is trademark of ArtiSoft Inc.
MS-DOS is copyright and trademark of Microsoft Corp.
Netware is trademark of Novell Inc.
PCBoard is copyright and trademark of Clark Development Company.
PKZIP and PKUNZIP are copyright and trademark of PKWARE, Inc.
ZDCS is copyright (C) 1991, Michael W. Cocke.