pete@Octopus.COM (Pete Holzmann) (11/07/89)
On the net, there is a long tradiition of distributing significant software
packages in source code form. Some might argue the reasons behind this fact,
but it is a fact nonetheless.
The entire net has benefitted from this. Most particularly, those who are
interested in furthering this worldwide computer communications network have
been able to build on early developments, creating ever more powerful and
easy to use software packages.
Recently, I've seen what I consider to be a dangerous threat to this.
Several people have taken the information made available through others'
hard work (i.e. the UUCP specs, the DCP/UUPC code, etc), added some
(significant, to be sure) improvements, and then released the results as
shareware.
I am not against shareware. Shareware is one valid way for people to eat
and program, when they wouldn't otherwise have a way to pay for the groceries
and write software at the same time.
However, it is a Bad Thing to stifle the ongoing releases of new and better
software in source code form. Also, shareware authors who release nothing
of value in source form are taking something significant from the net (our
collective network mechanism for widespread software distribution) without
contributing anything of value to everyone. Yes, they make their software
available to us, but (1) we must pay for it, and (2) we *can* place it on
the net ourselves.
I don't want to get into an argument about shareware here. I *do* want to
make a positive proposal that will help the net as a whole, and will
improve the acceptability of shareware on the net.
My proposal is this:
Shareware authors should choose a *significant* portion of their algorithm
and/or code, and should release it to the net in source code form. By
significant, I mean that they should choose something that hasn't already
been done on the net.
This will not harm a good shareware author in any way. Sure, you're giving
away a few of your "secrets", but (1) you aren't giving them all away;
(2) the package you sell is far more than just an algorithm or three! In fact,
the quality documentation, user interface and installation procedures are
usually far more work and added-value than the lower level code.
It should be noted that 'arc' (bad example I know :-)) is available in
complete source form, and yet is still sold as shareware.
The benefit to the net is that others can continue to build incremental
improvements, benefitting from your knowledge, just as you have benefitted
from others in the past.
Just as an example: there has been a rash of news/mail-for-PC shareware
postings recently. All are based on the uucp-for-DOS work done a year or
two ago. C'mon folks, let's see some source code so somebody else can
improve on YOUR work! For example, several people have made improvements
to the uucp transport code (uuio). Let's collect those improvements and
put together a really good uucp protocol driver, based on everybody's
knowledge, rather than reinventing the wheel every time!
The time has come to cooperate more.
Pete
--
Peter Holzmann, Octopus Enterprises |(if you're a techie Christian & are
19611 La Mar Ct., Cupertino, CA 95014 |interested in helping w/ the Great
UUCP: {hpda,pyramid}!octopus!pete |Commission, email dsa-contact@octopus)
DSA office ans mach=408/996-7746;Work (SLP) voice=408/985-7400,FAX=408/985-0859dar@nucleus.UUCP (Dario Alcocer) (11/09/89)
> My proposal is this: > > Shareware authors should choose a *significant* portion of their algorithm > and/or code, and should release it to the net in source code form. By > significant, I mean that they should choose something that hasn't already > been done on the net. > > [ text deleted] > > The benefit to the net is that others can continue to build incremental > improvements, benefitting from your knowledge, just as you have benefitted > from others in the past. > > [ text deleted] > > The time has come to cooperate more. > I am relatively new to the net, but it seems to me, based on past experiences, that people in the unix world (especially those in the academic community) seem more open to the development of a collective body of knowledge so that everyone can prosper. In contrast, I see most people involved in the PC world (with the exception of some writers in DDJ, PC-Mag, etc.) generating software for the commercial environment, and seem less interested in sharing knowledge that in making a buck. I know that this is a generalization, but I think it holds true. Here at work, I'm suprised how many times we have to 'reinvent the wheel' because PC developers are concerned about others making money with source code or algorithms they developed. If I have offended anyone, my apologies, I just wanted to share my general impressions. However, I remain open to any other points of view. Dario Alcocer (San Diego, CA) via Nucleus (dar@nucleus.mi.org)