Re: Funding software development?

On Thu, 13 Jun 2002, Francois Gouget wrote:

> > Certainly there would always be some enclaves of secret code, but maybe
> > less than you think.  The Navy has already built at least one "smart" ship
> > (that ironically ended up dead in the water) based on Windows NT.
> 
> And which proportion of code did NT represent? I am sure it really was
> not that much. So I think it is you who are overestimating it. But of
> course that's just my opinion against yours.

The proportions are irrelevant.  The point is that mission-critical code 
used for a military purpose was incorporating off-the-shelf commercial 
software.  If they would do that, why wouldn't they use GPL software at 
least as easily, if not moreso?  GPL software could then dominate in the 
military markets as readily as others.  That was my point.

> > It's not necessary to drive other licenses and truly free software to
> > extinction.  It's only necessary to drive them to irrelevance.  That's a
> > very different matter.
> 
> And what do you mean by 'irrelevance'? And how is that a bad thing if
> there are still numerous widely used project using such licenses?

What do you THINK "irrelevance" means?  If the GPL grows sufficiently 
powerful that other licenses no longer matter...  it should be obvious.

> > Rightly so?  Is it really Transgaming's fault if the users aren't 
> > willing to support development?
> 
> No. But their handling of the situation is their doing.

I agree they could have handled the situation better.  However, nothing 
I've seen has shown evidence of actual wrongdoing, as others have claimed.

> > From what I've seen, it appears that if anyone has the right to be bitter,
> > it's Transgaming and its too-few subscribers, not the whole Wine community.
> 
> Yes they may be bitter that they did not have the number of subscribers
> that they expected but I don't see how that's the Wine community's fault
> since it defacto went as far as to not give them any competition.

Let's be consistent here.  Either their development efforts were worthwhile 
(and the Wine users should have supported their efforts be subscribing) or 
they weren't (in which case the Wine developers shouldn't have restrained 
their competitive efforts).

The events here suggest a mixed message -- the developers acted as if the 
coding by Transgaming was worthwhile, but not enough users seemed to agree.

> > > The way I read their original statement is that they would keep the
> > > Direct3D code until their subscription quota was met but that the rest
> > > of the code, e.g. COM, would be returned to Wine. Yes it is now obvious
> > > that we (Me, CodeWeavers and the Wine community) were naive but then we
> > > (the Wine community) learnt and switched to the LGPL.
> >
> > So it's Transgaming's fault that you misread their statement?
> 
> Did I? Again, I don't think so. Did I mis-read Ove's email? I don't
> think so either.

I'm not sure offhand which email you're referring to, but if it's the one 
where he asked for help with something so "we" could get to work on another 
thing, he didn't specify who "we" meant, so it could as easily mean the 
company, not the community.  (And given a company's business orientation, 
it's best to assume they mean the company by default.)

> > If their business model fails, so be it.  However, it's one of those models
> > that sounds great on paper, which Open Source advocates have encouraged
> > companies to pursue.  Now, a company tries to pursue it, and gets attacked
> > by the community in the end.  Why?  Because not enough people wanted to
> > support the model to make it viable.  Yet, despite this lack of support
> > from the community, the community nevertheless demands that Transgaming
> > should still follow through on their end of a bargain never executed?
> 
> I don't. I would be quite happy if Transgaming apologized to their
> subscribers and declared that their Direct3D implementation and their
> future developments are proprietary and will remain so for the forseable
> future. Add to that a license change from the AFPL to one that forbids
> redistribution of WineX binaries (which is effectively their intent)
> and I would be happy.

Well, that represents giving up on the model.  But maybe it's time.  (But 
unless they plan to refund money to subscribers, they should still fulfill 
their promise if the goal is ever met.)

> We would then have a clear situation where the Wine and ReWind community
> both know where they stand. This would not prevent them from making
> contributions to Wine or ReWind, but at least it would not create and
> nurture false expectations.

Let's give them the benefit of the doubt -- maybe they nurtured those false 
expectations only because they already held them?

> Now I think that with their recent statement their are making a step in
> the right direction.

I guess we'll just have to see what happens next.

> > This is certainly true when the project depends entirely on volunteers,
> > unless those volunteers are so invested in seeing the results that they'll
> > endure unenjoyable work to achieve them...
> 
> If you follow this philosophy, good luck with Gangplank!

Huh?  I said that open-source projects SHOULD be enjoyable if they're run 
by unpaid volunteers, even if some volunteers are willing to endure it...

All of my efforts on Gangplank have been on a volunteer basis, because it 
was interesting work in its own right.  Without enough users, it ceased to 
be interesting enough to keep working on, and I ultimately released it as 
open source (under the QPL) hopeing to build a user base.  Unfortunately, 
I've still got some more effort I need to expend before it's ready for a 
stable 1.0.0 release, and I never seem to find the time...

But I never mentioned Gangplank in this conversation; what made you go 
searching for it?  (After all, it's certainly not a well-known project!)

I might point out that while I didn't use the GPL license, neither did I 
use the BSD license.  I chose the QPL license primarily because it helps to 
discourage code forks, and I didn't think I'd have an active enough project 
to naturally discourage forks, at least not at first.  Once the project is 
better established, I could imagine switching to a license like the MPL or 
GPL (more likely MPL), and if it becomes particularly well-established, 
maybe eventually BSD or X11.  For now, the QPL makes sense for this code...

Deven