Public Domain Responsibilities
By Richard Greenlaw
From Dr. Dobb's Journal, Number 81, July 1983, Page 10
To DDJ Readers and CP/M Users,
Over the years that DDJ has been serving us, I have
contributed on article (a disassembler in issue no. 27), a review
of the PC-1 pocket computer (in issue no. 52) and a few letters.
However, this contribution is a bit different. The bottom line
of this letter is a warning!
I am the author of a set of programs which is in
widespread public domain distribution (at least three
continents). They are called SQ.COM, USQ.COM and FLS.COM. They
run on CP/M systems and serve to squeeze files to smaller sizes
in preparation for transmission via modem or distribution
diskette.
The latest version of these programs are 1.5, 1.5 and 1.1
respectively. I released these programs to the public domain,
except for financial gain, and included full source code (in C
language) and extensive documentation. This software is now
available via the C Users group, which has advertized in DDJ.
Having had reasonable experiences with previous public
offerings, I gave my name and address in the documentation and
in the sign-on message of each program. Therein lies the story!
In recent months I have had a couple of dozen phone calls
claiming that these programs have destroyed various innocent
files. But when I ask for details, I find that the programs in
question have version numbers higher than any I have ever issued!
Not only that, but many different version numbers, as high as
1.9, have been mentioned!
I expect that whoever produced each version of these
programs was as innocent as I was in August of 1981 when I ussued
the above-mentioned programs. But the fact remains that one or
more persons has taken it upon themselves to propagate modified
versions of my software without taking the responsibility of
adding their name and address to the documentation and sign-on
messages.
It is an unfortunate fact that C and probably other high-
level languages can crash impolitely when the CP/M TPA is too
small for the program and its data space, including dynamically
allocated local variables. (These programs do use recursive
algorithems).
I have heard, without confirmation, that vast amounts of
code have been added to some versions of these programs to add
wild card file name processing. Perhaps this is the problem. I
provided that feature as a separate program, FLS.COM, precisely
for that reason. It is also possible that full and careful use
of the recompilation options available in BDS-C has been abused,
resulting in too much code or improperly linked code. Or maybe
there have simply been some accidents in copying or modem
transmission.
Whatever the case, I recommend that if you modify
software which you have not written, then try to consult with the
original author before publicly distributing such software. If
that is not possible, at least accept the responsibility of being
the first line of complaint handling by adding (not replacing)
your name to the documentation and sign-on message.
If anyone has a file which the above-mentioned versions
of my programs cannot squeeze and then reproduce, please contact
me. My computers can read and write the following diskette
formats: Micropolis Mod II, Max-80 5" 40 track/side (SS/DS,
SD/DD) formats, Osborne Single Density, Omikron 5" single
density, Xerox 5" single density. Except for the Micropolis, all
of the above are handled by my Lobo Max-80 computer.
Do you really want more public domain source code
sofrware? If so, and if you know the source of these
modifications, please help stamp out the bad versions (copies?)
of this software. Unfortunately, I don't know exactly which
versions are bad because not one single person has ever sent me a
courtesy copy of any modifications!
To those bewildered innocent victims, the unsqueezing
precess contains a validity check based on the original file. If
you can produce a file (not just a console display) by running
the USQ.COM program, then the resulting file should be an exact
replica of the original.
I also ask purchasers of communications packages who I
have authorized to distribute this software not to assume that I
am responsible for the communications packages. It seems that
reproduction of my documentation files makes my name and address
more prominent than those of the commercial distributers!
Sincerely,
Richard Greenlaw
251 Colony Ct.
Gahanna, Ohio 43230
