GNU A Possible Ally?

Jordan Henderson jordan@Starbase.NeoSoft.COM
Sun, 11 May 1997 10:47:22 -0500 (CDT)

> From: Kelly Murray <>
> I wasn't going to follow up on this GNU stuff, but this isn't
> quite the same topic, so please excuse me.
> >From: Jordan Henderson <jordan@Starbase.NeoSoft.COM>
> > How about this scenario.  LispOS develops to some level of maturity.
> > All of the CL Vendors grab a (fairly) stable snapshot and bolt on
> > their own CL technology.  They then put their people to work on
> > incompatible extensions.  There is great value in each supported
> > system and many of them have advantages over the PD LispOS, one of
> > which being a commercial firm that takes responsibility for 
> > maintenance.  Users select one of the many LispOS variants for
> > their own work.  The impetus to improve the PD LispOS is much diluted
> > because the much of the large potential user base is off using commercial 
> > variants, which all move in their own various directions. 
> >
> > This, I think, is a realistic scenario.  The only way the LispOS could
> > gain some mindshare in the world at large is if there was one product 
> > that would steadily be improved to meet the needs of users.  Anything
> > else is marginilization.
> This is pretty much what happened with UNIX, scores of modified
> proprietary versions.  A huge commercial success.

Someone (Mr. Berger?) has already pointed out what's happening to UNIX
in the marketplace right now.  I would say it's being marginalized and
it would have been a much stronger competitor had real progress been
made in the 15 years that Unix was a commercial competitor.  Instead,
we had a bunch of interests redoing the work, over and over again
necessary to make Unix a stable and user-friendly system.

We know that Unix is not being marginalized due to superior technology.

Also, note that the original licensor of UNIX technology AT&T never
made money on this technology.  Then, Novell couldn't make money on

> It seems to me this objection, which I believe was also a motivation
> for RMS to create GPL (commercial emacs more popular than his free
> one), shows the fundamental flaw in the why GPL impedes progress.
> GPL'ers don't want the commercial competition,
> because they know they can't keep up.  
> So they don't allow it.

Well, the commercial emacs's haven't "kept up".  Please, before you 
make universal generalizations make sure you don't also give the 
example that disproves it.  It looks rather silly.  I guess emacs is
"the exception that proves the rule", huh?  (That's a silly phrase,
the one thing an exception NEVER does is prove a rule, it just says
the rule isn't entirely general, like all rules.)

> Competition is a very good thing.  Cooperation is also a good thing.
> You need both to make progress.  

Now, I agree with this.  There are a lot of things that competitors do
better than cooperators that do make a better product.  For example, if
it hadn't been for the Red Hats, et al, Linux installations would have
been something for hackers only.  I am willing to listen to licensing 
ideas that won't mean that the users aren't all attracted to the 
commercial implementations, marginalizing the LispOS that I would 
like to use (and be able to modify), but I haven't heard one.

How about we form an organization that holds the copyright on the original
material and license it to commercial organizations for their use (for 3-5
years at a time) while still giving out the stuff to hobbyists, researchers,
educators, etc.  We could use the money for supporting development that
would be free (for non-commercial use) and when the time came up for the 
commercial organizations to relicense we could trade the licensing cost 
for parts of their system that we find particularly attractive to put 
into the public domain.  The licensing costs could be based on # of deployed 
systems, indexed to inflation based on the original license cost.  The 
commercial organization could choose to pay in cash if they don't want to 
trade technology.  Maybe all of their interface docs based on derived works 
would have to be public domain too, so we could cleanroom redo the really 
attractive parts.  Any problems with this?

The base problem with making it PD is that the commercial organizations will
benefit from FREE software and do NOTHING to benefit PD software.  I do see
this as somewhat unfair.

> -Kelly Edward Murray