home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Hall of Fame
/
HallofFameCDROM.cdr
/
util2
/
wssi330.lzh
/
WSSI300.DOC
next >
Wrap
Text File
|
1986-09-06
|
53KB
|
1,123 lines
Instructions for WSSINDEX version 3.00
Introduction
Have you ever had to search through dozens of disks trying to
find a seldom used program which you knew you had somewhere?
Well here's the solution. WSSINDEX.EXE is a program for keeping
track of MS-DOS files stored on floppy or hard disks. A database
is created which contains, for all of your disks, the information
normally available from the DOS DIR command, as well as optional
comments and categories describing each file. The information in
the database may be printed with various sorting options, or may
be searched for specific file names with wild cards in case you
don't remember the exact name of the file you are looking for.
Additional functions include listing files without current backup
copies and extracting directory information from ARC files. The
program is menu driven and easy to use without referring to the
documentation, but please read it once anyway.
Program operation
From the operating system prompt, type WSSINDEX (there are no
parameters). The screen should clear, a copyright and version
number message and a "Do you want instructions" message will be
displayed. Here, and many other places in the program, a default
option is displayed which will be used if you just hit return.
If you reply Y (just hit return), you will get an abbreviated
listing of this document on screen. Next the program searches
for a configuration file called WSSINDEX.CNF on the logged disk,
or initializes with default options if it is not found. If a
configuration file is not found, the default configuration is
displayed and you are given the option of reading a configuration
file or changing the configuration. You may also change the
configuration information later with the "O" and "E" options.
You may change the default name of the configuration file by
using the DOS SET command to assign the desired file name to an
environment variable WSSICNF. The syntax is
SET WSSICNF=filespec
where filespec is the complete name of the configuration file,
including drive and subdirectory if necessary. (Actually, if
filespec is not a complete file name, omitted parts are filled in
from "WSSINDEX.CNF" as described under the "G" option.) You may
want to put this command in your AUTOEXEC.BAT file. It shouldn't
matter whether the filespec is upper or lower case. You can
review the environment variables by typing SET at the DOS command
prompt. If this environment variable is defined, it is assumed
that you are an experienced user, and the "Do you want
instructions" question is not asked; you proceed immediately to
the main menu (assuming that the configuration file is read
successfully).
In either case, after configuring, the main menu is displayed.
The menu options are:
A - Add (or update) disks to the database
Insert the first disk to be indexed into the indexing drive
and hit the space bar when ready. The processing time is
roughly proportional to the number of files in the directory,
and will typically be only a few seconds for a floppy disk.
If the disk is not labeled, you will be given the option to
either write a label on the disk, enter an 11 character
string to be used in place of a label, or abort back to the
main menu. See the "L" option below for a discussion of
"legal" volume names. You will be informed if the volume
name matches a volume already in the database, and given the
option to replace the previous entry or not enter the disk
into the database. The informative message "NO FILES IN
d:\*.*" will be generated if there are no files in the root
directory, along with similar messages for other empty
directories. When prompted you may either insert another
disk and hit the space bar, or hit any other key to return to
the main menu.
An important feature: if you replace a disk in the database,
any comments or categories which have been previously entered
are not lost, they are copied over to the new entry if the
file and subdirectory names match. Thus you should feel free
to re-enter a disk into the database after even small changes
have been made. During a replacement operation, the old and
new entries are temporarily in core at the same time, and
both count toward the limits on the number of files, disks
and subdirectories.
There is one circumstance which could make replacing a disk
in the database very time consuming. The code which copies
old comments to new entries is a straightforward doubly
nested loop, which means that the running time is roughly
proportional to the square of the number of commented files.
This is not a problem with floppy disks which seldom have
even 100 files, but on a hard disk with a few thousand files,
the processing time could stretch to several minutes. The
obvious solution is to sort both sets of names first, which
would reduce the processing time for commented disks, at the
expense of increasing it for non-commented ones, and I may
implement this in a future release. Comments from registered
users are solicited.
B - list files which are not Backed up
This options is intended primarily for listing files on a
hard disk which have not been backed up, but can also be used
on floppies. A file is considered to be backed up if another
file with the same name and the same or later time tag is
found on another disk (possibly in an ARC file). Exact time
and size matches are not required because some copy programs
do not preserve the old date, or even the size for files
which are not padded to a full disk sector.
When you select this option, the database is first sorted
using an internally defined set of sort keys. Then you are
given the usual selection options (described under the "P"
option), including disk name, but it doesn't make much sense
to select this way since you are only listing files on one
disk anyway. For the same reason, you might as well set the
configuration to skip printing the disk name (it will be
listed in the title anyway). You will almost certainly want
to configure the exception list to exclude *.BAK (or whatever
your favorite word processing program names its backup
files). And don't forget to request use of the exception
list, otherwise you will have to stop the listing with
control-c and try again.
If you have a hard disk which is logically partitioned so
that it looks like multiple disks, you probably don't want to
consider a file to be backed up just because it appears in
two partitions. To avoid this problem you can either load a
database with only floppies included, then index a single
partition before checking it for backups, or you can use the
"R" option to remove all but one partition from your
database.
C - add or modify Comments and categories for files
See below for a description of the add-comments menu. Files
to be commented may be limited by the usual selection options
(described below under the "P" option).
D - Display statistics
This option displays the number of files, subdirectories and
disks in the database, and the amount of contiguous core
remaining for additional entries. There will always be at
least one subdirectory listed, since the root directories of
all disks are treated as one subdirectory. Some points about
the reported core remaining: it is only the contiguous space
remaining which is reported. If a disk is removed or
replaced, a "hole" is usually created in the memory space
which will be used for the next disk added to the database.
Some operations may create fragments of core too small to be
useful, hence the space remaining may change slightly if you
save the database on disk and then reload it. You will
receive a warning if the available space drops below 4K. The
program may not always survive running out of core, so it is
best to do a save to disk if you get this warning. Also note
that opening a file requires at least 2K for buffers, so you
may not even be able to save your database if there is too
little core free.
E - change prEdefined categories
This option displays the 26 predefined categories available
in the add-comments section (or anywhere else that multi-
character text input is allowed) and allows you to change
them. These categories are stored in the configuration file,
not the database file.
F - Find a file in the database
Find is equivalent to Print (described below) to the screen
without resorting.
G - Get a database from disk
Enter the name of a file previously saved with the "W" menu
option. The default name is the first of
1. The last file name used in a Get or Write command
2. The DOS environment variable WSSIDIR (DOS command
set WSSIDIR=filespec)
3. WSSINDEX.DIR
which is defined, but you may override all or part of the
name if desired, including specifying a drive letter or path.
For example, if you enter C:\JUNKDIR\.LST, when the default
is WSSINDEX.DIR, the database will be loaded from
subdirectory JUNKDIR, file WSSINDEX.LST on drive C. You will
be warned and given an opportunity to escape if you have a
database in core which has been changed without being saved.
Otherwise, the only way of aborting this option is to enter a
non-existent, or better illegal, file name such as T.T.T.
I - change Indexing drive
Enter the disk letter to be used for subsequent indexing.
This does not change the logged drive.
L - Label a disk on the indexing drive
The current volume label, if any, is displayed and you are
prompted for a new label. A carriage return aborts back to
the main menu. The labeling is done by a DOS call, so DOS
restrictions on allowed characters apply (in particular, your
input will be forced to upper case because lower case letters
are not allowed, but blanks are ok). If in doubt, try the
label you want to use; an illegal name will be rejected with
no harmful consequences. I would suggest that you not trust
a copy protected master disk to any disk labeling program;
who knows what sort of games they might have played with the
directory.
Note: if you have a disk with a mixed case label (which you
can easily create with DEBUG or some other utilities),
WSSINDEX will not be able to change the label (DOS rename
call fails), the reName option will not work, and the Remove
option will only work if the disk name is read from the disk
rather than the keyboard. Disks with labels differing only
in case are considered to be different when adding disks to
the database, but cannot be distinguished when selecting by
volume name. If you use really weird volume labels, such as
ANSI cursor positioning commands, you probably deserve
whatever happens to you.
N - reName a disk in the database
You are prompted for old and new names and the disk name in
the database is changed. A carriage return aborts back to
the main menu.
O - change cOnfiguration info
This option allows you to set, save or restore the same
options as the initial configuration display (other
configurable items are set with the "E" option but also saved
and restored here). Default values are displayed after each
prompt, so to make a small change in the configuration you
can just hit the return key for most items. The options are:
1. Indexing drive mode letter.
2. Maximum number of files, disks and subdirectories. This
information cannot be changed with a database in core.
These numbers cannot be larger than 16,379, no matter
how much memory you have. With a 256K machine, you will
probably be able to accommodate around 3000 files; with
640K you may be able to hit the 16K limit. Note,
however, that a 16K file database will probably be too
big to fit on a 360K floppy disk. These numbers assume
that you don't enter very many comments.
3. Extract information from .ARC files (answer yes or no).
See below for a detailed description of .ARC file
processing.
4. Create .BAK files if saved file name matches an existing
file (answer yes or no). Normally, a file with a
matching name would be renamed with a .BAK extension,
and an existing .BAK file would be erased, but you may
elect to overwrite instead (with verification). This
option is provided for users with limited capacity
floppy disk drives who may not be able to fit 2 copies
of a database on the same disk. It is obviously
dangerous to reply yes as even a failing write operation
will probably destroy the old file.
5. Printer width and length. This information is used to
decide when to print new headings and whether or not to
start a new line for comment and category information
(even if "printing" to disk). It is assumed that your
printer is at least wide enough to hold all printed
fields except comments and categories on a single line.
This may not be true if you have subdirectories nested
deeply enough, in which case you will have to suppress
selected fields when printing. The program does not add
any blank lines for top and bottom margins, so you may
want to adjust the number of lines specification
accordingly.
6. Print comments and categories on the same line as
directory info (if they fit), yes or no.
7. Pause at end of page? For those of you who use single
sheet printers. If you reply no, each page after the
first will begin with a form feed (even if "printing" to
disk).
8. Select information to be printed. File size, date,
time, seconds field of update time, volume name,
subdirectory name, name of .ARC file, comment or
category may be suppressed for option "P" output. Note
that the time field is automatically suppressed if the
date is omitted.
9. Start in expert mode (menu display off). You will
probably want to say yes after you have used the program
for a while.
10. Exception list. A list of up to 10 special case names
which can be excluded from printing. By default, this
list consists of only COMMAND.COM, but you might also
want to add CONFIG.SYS, AUTOEXEC.BAT or anything else
you have many copies of. Wild card names are allowed.
If you read or save a configuration file, the default file
name is the first of
1. The last configuration file saved or read
2. The DOS environment variable WSSICNF (DOS command
set WSSICNF=filespec)
3. WSSINDEX.CNF
which is defined.
The format of the configuration file often changes from
version to version; compatibility is not guaranteed.
However, I do try to arrange things so that most information
can be extracted from an older configuration file before a
read error occurs. One of the first things you should do
with a new version is to review your configurations settings.
P - Print (and sort)
Files are optionally sorted in ascending or descending order
by any combination of
1. file name 6. date and time
2. extension 7. category
3. disk name 8. comment
4. subdirectory 9. .ARC file name
5. size 10. .ARC file extension
When prompted, enter a comma-separated list of sort key
numbers, changing the sign of any you want in descending
order. Most of the keys are self-explanatory, but the last
two need a little explanation. If you have elected to
extract information from ARC files, the file name and
extension used with keys 1 and 2 for members of an ARC file
are those of the ARC file member. Thus, with only keys 1 and
2, the ARC file members sort by their own names. Keys 9 and
10 select the ARC file name for members of an ARC file, or
the usual name for ordinary files. In addition, there are
two special conditions checked:
1. If sort keys 9 and 10 are used and have higher priority
than 1 and 2, the ARC file itself will sort ahead of any
of its members;
2. If keys 1 and 2 have higher priority than 9 and 10, then
ARC members (and other files) are sorted by their own
names.
An interleaved list of keys such as 1,9,2,10 will probably
not do anything useful. If there are multiple copies of an
ARC file, you will probably want to have the disk and/or
subdirectory as an intermediate key when sorting first by ARC
file name. To summarize, 9,10,3,4,1,2 will list ARC members
following the containing ARC file, while 1,2 will list all
files alphabetically.
If you elect not to sort (sort key equal zero), the current
sort order is preserved. After sorting, files to be listed
may be selected by
1. name and extension 5. date range
2. subdirectory 6. size range
3. disk name 7. comment
4. category or 8. excluded
if in a list of configured exceptions. Selection option
numbers are entered in a comma-delimited list; the order is
not significant. All but option 8 lead to prompts for more
information. For comment (category), enter a wildcard
string, or * to match any existing string, or just a return
to match absence of comment (category). Printer information
in the configuration file is used to control paging.
Information may be "printed" to printer, screen or disk. Disk
output might be massaged by a text editor to change the
format, or transported to another machine for printing; it is
not readable by the "G" menu option. On most systems, you
may abort printing by hitting control-C, but I recommend that
you test this feature before you count on it. Control-s
should make any screen display pause until you hit another
key (control-q may be required), but this is an operating
system feature not under the control of WSSINDEX.
Q - Quit and return to DOS
You will be warned if the in-core database has been changed
without being saved, and in any case you must verify that you
really want to exit. For your protection, the default option
is N (don't quit) if the in-core database has not been saved.
R - Remove a disk from the database
The volume name to remove may be either typed in or read from
disk.
S - Sort without printing
The purpose of this option is to set the order in which files
are processed by the add-comments option. If identical files
on different disks are commented sequentially, the copy
comment function is often useful. The usual sorting options
(listed under the "P" option) are available. The sort is
done using the Quicksort algorithm. The sorting time for a
randomly ordered file is O(N*logN), typically 30 seconds for
2000 files. If, however, you should somehow manage to enter
your files in alphabetical order, you will discover one of
the quirks of Quicksort: the sorting time for an ordered file
is O(N*N), turning a 30 second sort into a 10 minute sort.
For this reason the database is stored unsorted.
V - list Volume names
The volumes in the database are listed to printer, screen or
disk, sorted by name or by decreasing free space. The
printer configuration information is used to control paging.
The display for each volume includes the volume name, the
number of free bytes, the total number of bytes available,
the number of files (excluding hidden files), the number of
subdirectories other than the root directory, the date the
disk was last added to the database, and whether or not the
disk is bootable. The disk is assumed bootable if it has a
COMMAND.COM file, the presence of the hidden system files is
not verified. On most systems, you may abort printing by
hitting control-c, but I recommend that you test this feature
before counting on it.
W - Write database on disk
The in-core database is saved on disk. The default file name
is the first of
1. The last file name used in a Get or Write command
2. The DOS environment variable WSSIDIR (DOS command
set WSSIDIR=filespec)
3. WSSINDEX.DIR
which is defined, but may be overridden in the same manner as
described under the "G" option. A previously existing file
with the same name will either be renamed with a .BAK
extension or overwritten (with verification), depending on
which configuration option you have selected. If creation of
.BAK files is enabled, you will not be allowed to use a
filename with a .BAK extension.
X - toggle eXpert mode
Start or stop displaying the verbose menu before each command
prompt.
.ARC file support
ARC is a shareware program by Systems Enhancement Associates (21
New Street, Wayne, NJ 07470) which maintains libraries of files
in special files with .ARC extensions. Files stored by ARC are
compressed by the one of several possible algorithms which
produces the smallest output file. ARC appears, at least in the
Boston area, to be supplanting the LU and SQ/UNSQ family of
programs which separately perform the library generation or
compression functions. Starting with version 3.0, WSSINDEX has
the (optional) capability of extracting information from an ARC
file, much like it does from subdirectories. Among other
advantages, this means that you can keep backup copies in ARC
files and cut down on the number of floppy disks needed to back
up your hard disk, and still be able to use the "B" option to
check for current backups. The only function which is not
allowed for a member of an ARC library is "view-file" since
WSSINDEX does not know how to uncompress the file.
ARC files introduce a complication in the operation of replacing
a commented disk in the database. What should be done if a
member of an ARC file matches the name of an ordinary, commented
file in the same subdirectory? Quite probably the files are
identical, but they could be completely different, and WSSINDEX
has no reliable way to tell which is the case. The current
program version uses the following algorithm to decide whether to
copy a comment after it has found matching file and subdirectory
names: if both files are ARC members, the ARC file names must
match, and if only one of the pair is in an ARC file, the size
and date must match. I am not completely satisfied with this
algorithm and am open to any suggestions for a better technique.
Extraction of information from an ARC file is noticeably slower
than reading a directory because the information is scattered
throughout the ARC file, rather than being concentrated in one
place. You will probably notice a lot of disk activity when
extracting ARC information.
WSSINDEX should not be affected by changes in future versions of
ARC (latest version I know of is 5.12), since these are likely to
involve only the encoding schemes, and not the header format.
There is, however, the possibility of having files with an ARC
extension which were not created by ARC. For example, the
Computer Innovations C compiler which I use stores its source
libraries in ARC files which are maintained by a program called
ARCH, and have a completely different format from SEA ARC files.
Usually, such files will be detected when the header is read, and
an informative message will be given. However, if the file
starts like a SEA ARC file, but later diverges, either because it
isn't one, or because it has somehow been mangled, WSSINDEX will
stop processing the directory of the disk. If this happens,
there will probably be garbage entries in the database for the
disk. In this case, your only options are to rename the
offending file, or to turn off the ARC extraction option.
Wild card names
Most sections of the program which require names also accept *
and ? wild cards. These are almost the same as used in DOS,
except that in WSSINDEX something like ABC* will only match a
file with no extension, whereas in DOS it matches any extension.
All comparisons are case insensitive, so lower case categories
(or even volume labels or file names if you manage to create such
non-standard beasts) can be matched by upper case inputs. Note
that if you have, say, a disk labeled WORD*, and another labeled
WORDSTAR, the string WORD* will match both.
The add-comments menu
The add-comments menu screen displays each matching file name
along with the disk and subdirectory it is found on. The ten PF
keys, alt-letter keys, plus BACKSPACE, INSERT, DELETE, ENTER,
HOME, and the left/right cursor arrows are used to control this
part of the program. A help menu and three sets of information
are displayed:
Copy buffer - a comment/category pair which can be copied into
the new comment field.
Current file - the comment/category currently in the database for
the displayed file.
New comment/category - the comment and category currently being
entered.
The special functions available are:
PF1 - Copy comment: if the file has a comment, copy it to the
new comment field; otherwise, copy the comment in the
copy buffer. If you want to replace an existing comment
with one from the copy buffer, you must first delete the
old comment (PF6). Note that this does not actually add a
comment to the file until you hit PF5.
PF2 - Copy category: if the file has a category, copy it to the
new category field; otherwise, copy the category in the
copy buffer. As with PF1, this does not actually add a
category until PF5 is hit.
PF3 - Copy both: the same as hitting both PF1 and PF2.
PF4 - Skip to next: advance to the next file, without changing
the current comment and category fields.
PF5 - Add comment and advance: the new comment and category (if
any) are added to the current file and copied to the copy
buffer. Blank fields will not be copied and will not
change the record for the current file.
PF6 - Delete comment: any comment for the current file is
deleted immediately. If you have entered something in
the new comment field, and decide that you do not want to
have any comment associated with the file, just hit PF6
to delete the old comment, followed by PF4 to skip to the
next file without adding the contents of the new field to
the database. It is not necessary to clear the text
entered into the new comment field.
PF7 - Delete category: any category for the current file is
deleted immediately. Proceed as described under PF6 to
delete a category after having entered something into the
new category field.
PF8 - Change search direction: toggles the search direction for
the next matching file between forward and backward.
PF9 - View file: the indexing drive is searched for the current
file. The disk volume name must match (an unlabeled disk
will be accepted). If the file is found, it is displayed
on the bottom half of the screen. Non-printing
characters are translated to periods, so there is no
problem with viewing a non-ascii file. The high bit of
each character is cleared to make WordStar files
printable. Lines longer than the screen width are
truncated. This function will fail and claim that the
file could not be found if there is not enough free
memory to open a file.
PF10 - Return to the main menu.
BACKSPACE - Delete character to left of cursor.
DELETE - Delete character under the cursor (note that the
functions of backspace and delete at some points in the
program are controlled by the operating system and may
not be the same as here).
INSERT - Toggle between insert and overwrite mode.
ENTER - Toggle between comment and category.
Left/right cursor arrows - Move around in the comment or category
strings for editing.
Control-left/right cursor arrows - Move the cursor in word rather
than character steps.
HOME - Move the cursor to the beginning of the current field.
(Hitting return twice will position you to the end of
the current field.)
Alt-letter - Copy one of the 26 predefined categories into the
category field, overwriting anything previously entered
there. (The behavior here is slightly different than in
other parts of the program were the predefined strings
are appended to any previous input). Note that changing
the predefined categories (E-option from main menu) will
not change the categories which have already been
assigned to files; it is the actual string which is saved
for each file. Also, the predefined category strings are
stored in the configuration file, not the database.
When all files which match the selection criteria specified have
been processed, you normally return to the main menu. This would
make it inconvenient to apply corrections to a previously
commented database, or otherwise process several groups of files
selected by name. So, if the only selection criterion is the
filename, you will be prompted for a new filespec when all
matching files have been processed. At this point, you can also
just hit return to return to the main menu.
Comments and categories entered for files are limited to 64
characters, although if you use the category field as it is named
it will usually be a relatively short phrase. Some users may
want to use this field as a version number. While WSSINDEX will
allow any character except a backspace, tab, linefeed, carriage
return, delete, escape or a null to be used in a comment or
category, you are advised to remember that control characters may
have undesired effects on your printer. The category field is
printed in parenthesis after the comment.
The distribution package
You should have received the following files, possibly combined
in a .LBR or .ARC file:
WSSInnn.EXE - the executable program. n.nn is the version
number.
WSSInnn.DOC - this documentation file.
WSSMnnn.EXE - a utility for merging WSSINDEX databases
(version number may not match that of the
main program).
WSSMnnn.DOC - documentation for WSSMnnn.
WSSIFRM.DOC - a description of the disk format of a saved
database.
RAINBOW.DIF - addendum to this document for Rainbow users.
INVOICE.DOC - an order form and invoice in case you need
one.
WSSITIPS.nnn - tips for using WSSINDEX
CP2.COM,.DOC - a public domain cut and paste utility
Hard/software requirements
You must have MS- or PC-DOS level 2.0 or higher to use this
program because it uses operating system features not supported
by DOS-1. Your operating system must recognize INT 10 hex calls
to scroll the screen, and INT 16 hex to read the keyboard. It
must also support ANSI escape sequences for cursor positioning
and highlighting. This support may be an integral part of your
system (as is the case for the Sanyo 55x series without the video
board option), but is often provided by an installable device
driver called ANSI.SYS (ANSI55.SYS for the Sanyo 55x with the
video board). If your system does not support ANSI.SYS commands,
you will see the control sequences on the screen; they begin with
a left arrow and a [. If you find that you are having this
problem, check to see that your system disk has the ANSI.SYS file
and a file called CONFIG.SYS which includes the line
DEVICE=ANSI.SYS, and possibly some other specifications. The
only MS-DOS machine which I know to have problems with WSSINDEX
is the DEC Rainbow for which a special version (3.00R) is
available. There is also a special version (3.00V) for the Sanyo
55x with the optional video board which utilizes the unused video
RAM at address F0000 to increase the storage capacity.
If you have problems
First, read the above section on hardware and software
requirements. The most likely source of trouble is missing ANSI
support, which really messes up the screen display. This program
is written in C. C library routines are notorious for producing
minimal error messages (such as FREE, ALLOC, WRITE) when
something goes wrong. Such errors should not occur in normal
operation, but the program may not always survive memory
overflow. Watch for the low core warning when adding disks or
comments to the database and do a disk save before continuing.
And now the commercial
If you would like to encourage continued development of low cost
software such as this, you may register by sending $25 (or any
amount you consider appropriate) to
Robert. W. Babcock
WSS division of DDC
4 Reeves Road
Bedford, MA 01730
USA
617-275-1183
Registered users will be notified immediately if any bugs capable
of destroying a database are discovered.
I don't mind receiving phone calls, but you will probably only
get to leave a 30 second message before my answering machine cuts
you off, and I do respond to correspondence. Long distance phone
calls will generally be returned collect, and always from the
above number, so you might want to refuse to accept and then call
me again.
If you have many guilty consciences to assuage, you might
consider a site license for $100 which covers unlimited use of
WSSINDEX throughout your organization.
Source code is available for $400 (less registration fee if
paid). If anyone is interested, contact the DDC Retaliation
Dept. at the above address for conditions of sale. In addition
to WSSINDEX, DDC also distributes a simple shareware BASIC
program called MCWSSA which aids in producing multi-column output
on a dot matrix printer using WordStar ($5 contribution
requested, also available in Sanyo 555 BASIC). Public domain
offerings include SPELLIST which lists a SpellStar dictionary
(BASIC and C versions available), and PC-KLUTZO, an Othello
program able to crush most human opponents.
Upgrade policies:
1. No additional fee is requested for version upgrades or
bug fixes which you obtain from a bulletin board or
another user.
2. Non-registered users may register and receive the latest
revision by sending $25 and a formatted SSDD or DSDD disk
in a reusable mailer to the above address. Note that I
am only able to write IBM-PC format disks.
3. Previously registered users may receive an upgrade the
same way, but for a $5 fee. (Please add $1 if your check
is not drawn on a US bank, and $2 for postage outside of
North America). No charge if you are the first to report
a significant bug, or for your first upgrade after a
registration which was not accompanied by a disk request.
Please include the version number when reporting bugs. Reports
of incompatibility with particular brands of machine will always
be accepted even from unregistered users (I can hardly ask you to
pay money if the program doesn't work on your machine). I will
try to produce special versions for any reasonably common machine
which cannot run the standard version, but only if someone with
said machine volunteers to do the testing.
In any case, you are encouraged to pass copies of this code on to
other users, provided that this notice remains in the code. This
program is supplied without any warranty, whether you are
registered or not. This program may not be distributed for other
than a nominal media and handling fee. Bulletin board operators
are authorized to make this program available for down loading.
I have heard that WSSINDEX has been listed by some of the
companies which sell or rent public domain program libraries, and
has been included in public domain software packages bundled with
low cost PC's. Just remember that if you buy shareware this way,
the author does not get any royalties, and you may not hear about
bug corrections or program enhancements unless you register.
Bugs and enhancements in previous versions
Enhancements are listed under the first version in which they
appear, bugs (hopefully) under the last.
Versions 1.01 through 1.03
1. Various minor bugs and a fatal error (destroys database)
when removing or replacing disks. These versions were not
widely circulated, but if you have one of them, you should
upgrade.
Version 1.04
None reported.
Version 1.05
1. Added I option and warning on exit if file not saved
2. First attempt at a Rainbow version (1.05R). Add comments
section didn't work (different function keys and ANSI
support). Couldn't read back a saved database because read
routine didn't expect a 5 character version number.
Version 1.06
1. Changed disk save code to create .BAK files instead of
overwriting old file.
2. Rainbow version needs special handling for delete key. Some
copies of version 1.06R claim to be 1.05R.
3. Line counting and pause at end of page logic fail if there
are comments.
4. Core used for comments was not freed when zapping an in-core
database before Getting a database from disk.
Version 2.00
1. Major revision. Add-comments section completely redone,
added category field and view file option. Added
configuration file, file selection options when printing,
expanded sort options, interception of control-c, and made
number of files limit configurable (had been fixed at 3000).
2. Disk file format change; upward compatibility retained.
3. First edition of this .DOC file.
4. Cursor positioning was not always correct when advancing to
the next file in the add-comments menu, and if not corrected
(say by hitting CR twice), added comment could be lost.
5. Print-to-disk operations did not check for disk-full errors.
6. A disk removal or replacement operation following a sort
messed up the sort pointers (resorting corrected the
problem, and the database could be saved to disk without
problems).
Version 2.01 (and 2.01R and 2.01V)
1. Copy comment/category functions changed to copy from
existing comment/category rather than copy buffer to allow
easier updating of a previously commented file.
2. First upgraded Rainbow release (there never was a 2.00R).
3. Category information could be lower case, but the match
string for selections is forced to upper case and thus could
never match.
4. If configured to not create .BAK files, the program would
not allow you to overwrite an old file with the same name,
after asking if it was OK to do so (introduced in 2.00).
5. The seconds field of the file date has always been wrong.
Version 2.10 (and 2.10V, 2.10R was never distributed)
1. Added predefined categories and "B" option. An error will
be reported when reading old configuration files because the
predefined categories are absent, but all other information
is retrieved correctly. Made all string matches case
insensitive.
2. "B" option didn't always work, reported some backed up files
as not backed up.
Version 2.11 (and 2.11R and 2.11V)
1. Fixed "B" option.
2. View file option crashes with a FREE error when reading an
unlabeled disk.
Version 2.2
1. Changed format of configuration file, some options will be
assigned garbage values if you restore an old configuration
file. Added option to suppress selected fields when
printing. Added default options for most input prompts.
Fixed core left function for greater than 64K remaining.
Deleted display of configuration info on initial entry if a
configuration file is found. Made default limits on
database size dependent on free core available.
2. Get database option doesn't warn you if there is a changed
database already in core.
3. Errors in reading configuration files were not always
detected.
4. Printer line width logic sometimes failed and gave an extra
line if you printed comments but not categories.
Version 3.0
1. Added support for .ARC files, WSSICNF and WSSIDIR
environment variables, default file names changed by
appropriate input/output to file, word-tabbing and insert
mode indicator in add-comments menu, enhanced selecting for
comment and category. Shortened the .EXE file by removing
unused floating point routines.
Possible Future Enhancements
1. A utility routine to convert an ASCII input file to a WSSINDEX
database. This might merely be a BASIC or C program which has
the necessary code to write a properly formatted database, and
hooks for the user to add code to read a particular format.
* * * * * * * * * * * Limited Time Offer * * * * * * * * * * *
If you have a disk catalog maintained by another program which
you would like to convert to a WSSINDEX database, you can send
me any documentation you have about the disk storage format, a
sample database including a printout and any ascii disk output
the program can produce, and a copy of the indexing program
(if that is possible without violating any license
agreements), and I will attempt to produce a conversion
utility. No guarantees that I will be successful or quick,
but if I do produce a custom conversion utility, you should
feel more than the usual moral pressure to become a registered
user. My hope is that there are only a few popular competing
programs out there, and that I will be able to build up a
small library of conversion utilities which will meet most
needs. This offer is somewhat of an experiment, and whether
or not it continues depends on how well it works out.
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
2. I have occasionally been asked why I don't support color.
There are two reasons: (1) I don't have a color monitor for
testing, and (2) playing with colors could increase
compatibility problems without significantly increasing
program utility. But enough users have asked for it that I
will probably add color support in a future version.
Registered users are encouraged to suggest other enhancements,
and past contributors are thanked for their ideas.
