Re: Transgaming's promises mean little (was: Re: bias...)

On 5 Jun 2002 [email protected] wrote:

> Nor are they ever likely to do so, despite their promises! 

While this statement demonstrates that you don't trust Transgaming, do you 
actually have any direct evidence that Transgaming will fail to follow 
through on their promise if and when they reach 20,000 subscribers?  They 
said there's roughly 3,000 now -- sounds like it could take quite a while 
to get to 20,000...

> In addition to the promise made to the wine developers to eventually 
> release the Direct3D work, Transgaming also made the following public 
> commitment:
> 
> http://web.archive.org/web/20011123120809/http://www.transgaming.com/businessmodel.php
> 
> "One aspect of TransGaming's model is based on the Street Performer 
> Protocol. We are licensing some of our 3D code under the Aladdin Free 
> Public License (AFPL), which restricts certain forms of commercial 
> redistribution. Users may freely download and use the software, but will 
> be encouraged to subscribe to our subscription service. We will not 
> release that code under a less restrictive license (such as the Wine 
> license) unless and until we have a paying subscriber base of at least 
> 20,000 users. This means that our work will not be fully incorporated 
> into the main Wine source base before that point..." 

Have they actually RETRACTED this commitment?  Or just stopped saying it?

> Transgaming then proceeded to take subscriptions to WineX based on this 
> promise to eventually return the work to the wine community.  At least 
> some of their customers took out subscriptions because they felt they 
> were helping to fund the development of free software for the wine 
> project.  See:
> 
> http://slashdot.org/comments.pl?sid=33287&cid=3596846
> http://slashdot.org/comments.pl?sid=33287&cid=3598924

Maybe they were, maybe they weren't.  It's too early to tell.  They entered 
into it knowing that there was a possibility that the subscription goals 
might not be met, and the code might never therefore be released.  It's a 
risk they chose to take, whether or not they realized it.

> Now it appears that Transgaming has reneged.  The promises that were made 
> to both wine developers and their customers have quietly disappeared from 
> their website:
>
> http://www.transgaming.com/businessmodel.php

You're jumping to conclusions here.  Just because they no longer promote 
that commitment doesn't necessarily mean they've reneged on it.  Maybe they 
have, but you can't draw that conclusion solely from this change in their 
webpage.  You should have more solid evidence than that before accusing 
someone of bad faith...

> Now all we have is: 
> 
> "Our business objectives are quite simple: work with the most successful 
> game developers in the world towards multi-platform releases of the most 
> popular games; enrich the gaming experience for end-users by providing 
> games on those platforms they wish to play."

This is an interesting comparison -- after all the people bitching that 
Transgaming should find a new business model, this rewrite of their webpage 
suggests that they've got a new one they're going to try.  That doesn't 
necessarily mean that they've reneged on the old one, but it seems they're 
not convinced that the subscription thing will work out, so now they seem 
to be trying to partner with game developers.  What's wrong with that?

Who knows?  If that approach is successful, they may decide to release 
their code regardless of the subscriptions.  Or they may follow through and 
release the code after they reach 20,000 subscribers, even if they aren't 
really banking on that target anymore.

The fact that this webpage changed is interesting, but doesn't prove that 
they've reneged on their promises.

> Evidently their business objectives no longer include contributing to 
> Wine (or Rewind, for that matter), as neither are mentioned here.

The point of the business model is to define how their business will make 
money and survive as a company.  Giving away code may not be relevant.

You're jumping to conclusions again.

> Now, it is perfectly understandable for Transgaming to have not released 
> the code until their subscription target was met (I assume it hasn't 
> been, but of course only Transgaming would know that for certain).  It 
> was also reasonable for the wine developers to change the license for 
> their software once they realised that these unfulfilled promises were 
> harming the project.

Whether or not harm was caused is a matter of opinion.  Supposedly there 
are nowhere near 20,000 subscribers (more like 3,000), so if that's true, 
there's no reason why Transgaming should have released the code by now.

> However -- nothing excuses Transgaming from this act of bad faith towards 
> their subscribers!  If Transgaming had ever been sincere in its intention 
> to return the Direct3D work to the wine project, there was no reason to 
> retract this promise.

What act of bad faith?  Show me where this promise was actually retracted.

> After all, they don't have to fulfill it unless they reach their 
> subcription target, at which point presumably their development costs 
> would have been paid for.

How do you know they won't?

> An honest approach would be to have simply acknowledged that they were 
> never likely to reach 20,000 subscribers, and then either released the 
> code anyway as an act of goodwill, or suggest that although they were 
> still prepared to keep the promise, the code would probably never be 
> actually released due to a chronically insufficient number of supporters.

Yeah, it would be nice if they would confirm whether or not they still 
intend to keep their promise.  There's nothing to prove they don't, so why 
assume an act of bad faith?

> At that point, the wine developers could have decided how to proceed, 
> weighing the risk of a possibly wasteful reimplementation the Direct3D 
> work by themselves against the small chance that Transgaming would 
> eventually release their code under the (old) wine license.

They may eventually reach 20,000 -- but what if it takes 5-7 years?  Maybe 
the Wine developers should work on what they want to, whether or not it 
might (or might not) be release by Transgaming in the end.

> Instead we now have this surreptitious retraction of their promise to the 
> community of developers and users :-(  Personally, I do not believe that 
> Transgaming ever had any intention of releasing the Direct3D code -- I 
> think it was merely a marketing ploy designed to attract additional 
> subscribers, and possibly also to discourage competitive independent 
> development in these areas within the wine project.  Whether by accident 
> or by design, these policies have succeeded remarkably well in ensuring 
> that wine is not a competitive alternative platform for hosting Windows 
> games.

Again, I've seen no retraction.  From the original webpage, it seems like 
they were trying to implement one of the proposed models for free software 
development.  It seems that it's been less successful than they've hoped, 
so they're refocusing on another approach.  Have they reneged?  We really 
can't know unless they actually reach 20,000 subscribers, then fail to 
release the code.  Otherwise, it's just guessing.

Still, it would be nice if they would clarify this situation, since it 
obviously causes concern...

> Nor is this the only example of Transgaming's insincerity -- 
> 
> Transgaming provides some of their game-related code under the AFPL.  
> This is not a free software license, but it does allow non-commercial 
> redistribution in source or binary form, among other things.  When 
> software is released under a license, this is also a "promise" of sorts, 
> since it assures the user that there will be no copyright infringement if 
> the uses of the code fall within the permissions stated in the license.      

Be glad they release anything under the AFPL at all -- the alternative 
would be not to release the source at all until the subscription goal was 
met.  They would have been within their rights to do that -- obviously they 
preferred to make the source available, but wanted to do it in a way that 
preserved the incentive for users to subscribe instead of taking it for 
free just because they can.

> Unfortunately, it appears that Transgaming has no intention of allowing 
> anyone to make use of their software under the AFPL -- the rights that 
> were granted are illusory, and will be taken away as soon as someone 
> tries to exercise them.  See:
> 
> http://lists.debian.org/debian-devel/2002/debian-devel-200205/msg02823.html

Legally, they can't stop it.  They said that they would PREFER it not be 
made readily available in a Debian package, and would also prefer not to 
change the license to prevent that, but would if necessary.  Obviously, 
they wanted to preserve the incentive for users to subscribe -- the easier 
you make it for them to get it for free, the less incentive they have, so 
of course they'd rather not see a Debian package to make it a snap.

Legally, of course, they couldn't stop it, but the Debian packager decided 
to respect their wishes and not create the package.

> Note that the code that was to be packaged is *not* the same as WineX -- 
> it provides only a subset of the functionality of Transgaming's 
> proprietary version anyway (and no support or testing).  Despite this, 
> Transgaming has clearly stated that if anyone makes use of this "lite" 
> version of winex, they will probably change the license to make that 
> impossible.  See:
>
> http://lists.debian.org/debian-devel/2002/debian-devel-200205/msg02469.html

Whether or not it's complete, it's still reducing the incentive for anyone 
to subscribe.  Since they could have kept that code proprietary instead of 
releasing it under the AFPL, I don't really see the problem here.  If they 
didn't care, they'd release that code under the X11 license.  Since they 
used the AFPL, they were clearly trying to make the source reasonably 
available without damaging the user's incentive to subscribe.

What's so evil about that?  They didn't threaten to send out lawyers.  They 
asked that it not be done, and threatened to take future steps to keep it 
from happening with future versions.  Both legitimate.  If they wanted the 
code to be distributed far and wide, and easily packaged, they would have 
used the X11 license instead of the AFPL.

> As Thomas Bushnell points out, Transgaming's distribution terms are a 
> scam -- if any group like Debian tries to make it convenient for their 
> users to make use of the AFPLed code, Transgaming will probably change 
> the license to prevent it.

Well, is that such a surprise?  They're trying to put the source out there 
for the benefit of developers.  If a few intrepid users take the trouble to 
build from source, fine.  Most users won't.  But if you package it and make 
it easy for all users, then the users have less incentive to subscribe.  
They could have put a no-binary-packaging clause on the license in the 
first place, but didn't.  Must you try to subvert their business model just 
because you want an easy package of what they put out for developers more 
than users?  Is it really such a surprise that they might defend their 
revenue stream?

> Since there is no long-term prospect of being able to release updated 
> packages of the code due to this threat, it's not worth anyone's time and 
> effort to bother doing so in the first place.  Needless to say, Debian's 
> developers decided not to bother packaging the AFPLed code, and 
> Transgaming got what it wanted without changing the license.

What's the problem here?  If you really want the stuff free, convince 
20,000 of your friends to subscribe, so that they release the code,  If you 
only want to get it for free, knowing that wasn't the intent, then you're 
being pretty disrespectful of the time, money and effort that they've 
alreeady invested.  (With a high risk of failure, I might add.)

> This allows Transgaming to persist in maintaining the illusion that they 
> are some type of "open source" company, while being wholly proprietary in 
> practice.

The "Street Performer's Protocol" is a valid open-source model.  But it's a 
critical-mass thing -- a few won't be enough to support a business, you 
need a lot of subscribers to make it work.  So, they're holding the code 
hostage until that critical mass is reached.  Not terribly unreasonable, 
given that the code wouldn't exist without their efforts, and it's the only 
real leverage they have.

> Finally, there was Transgaming's decision to fork the wine project and 
> create "rewind" (http://rewind.sourceforge.net).  Although they (along 
> with many of the loudest non-contributors who post here) claim to support 
> an X-licensed branch, this claim is not backed up by any significant 
> contributions of code.  Rewind was established over two months ago, yet 
> most (if not all) of the code contributed so far appears to have been 
> lifted from the wine-patches mailing list.

Well, it's not a surprise that they want an X11 tree to exist if that's 
what they need to make their business model work.  That doesn't mean you 
should expect them to be making massive code contributions to the X11 fork 
unless and until they reach that 20,000 subscriber mark...

> I agree with Dmitry Timoshkov when he wrote:
> 
> "Transgaming wants to include patches made for LGPL'ed Wine into WineX 
> and as an intermediate step they incorporate those patches into ReWind 
> (just to create an illusion that they support an X11 licensed Wine 
> tree)."

Well, duh.  In the short term, that's why they need Rewind to exist.  In 
the long term, it's possible that they feel that X11 is a better license, 
but in the short term, they want the chance to receive contributions from 
Wine developers WHO ALLOW IT.  If the developers refuse, too bad -- they 
have to live with whatever's in the X11 tree if they can't convince them.

> The evidence so far supports this interpretation.

Sure, but it supports other interpretations as well.  You may be leaning 
toward this interpretation, because you evidently distrust Transgaming.  
You're jumping to conclusions.  You might still be right, but there are 
other possibilities that don't require anyone to be acting in bad faith.

> It should be obvious by now that "rewind" is not a free software 
> *project* in any meaningful sense -- it has no independent community of 
> developers, and no commercial support; it has no mailing lists or other 
> fora for discussion.

So what?  MANY free software projects have been abandoned by their authors 
without getting this sort of denigration.  Does that make them no longer a 
free software project?  Well, maybe not active.  But there's plenty of dead 
projects out there to go with the live ones...

> Instead, it consists of two Transgaming employees whose "contribution" 
> consists of applying other people's patches to the old X-licensed 
> codebase (and from there into WineX).

Hey, even that minimal effort takes some work.  And it gets harder as the 
trees diverge.  And others can still benefit from this free software, even 
if that's not the real reason they're doing it.

> In short, it's merely a temporary code repository that is used by 
> Transgaming to indirectly capture patches flowing in one direction from 
> the wine project into Transgaming's proprietary software hoard.

They could have dispensed with the pretense entirely.  They could have 
asked for X11 permission, skipped the integration work into Rewind, and 
just worked on integrating it with their own prioprietary codebase.  They 
deserve a LITTLE credit for bothering to integrate it twice for the public 
project, even if it's mostly a sham.

> If the people who claim to support the use of the X license had devoted 
> as much effort contributing code to rewind as they contribute in windy 
> rhetoric to the wine-license mailing list, rewind's codebase might 
> actually have some independent value.

Who knows?  Maybe that will happen.  If things get much more acrimonious, 
a community divide may be the result.  In that case, you may see part of 
the community move from Wine to Rewind over license issues.  I don't think 
anyone really WANTS a divided community, but it certainly could happen...

> However, I won't hold my breath waiting to see that happen. It seems that 
> the X license is always the best way to license code -- as long as it's 
> some other sucker who is doing it!

That's the nature of the X11/BSD licenses...

> Indeed, it should be obvious by now that rewind will never have any 
> independent value, since it will always provide a (probably broken) 
> subset of the functionality available in either LGPLed Wine or 
> Transgaming's proprietary WineX.

It's a bit early to predict.  Other proprietary companies may pick it up 
and help improve it for all you know.

> The nature of the X license practically guarantees this.  Therefore, it 
> would be simpler for all concerned to simply dispense with this charade 
> -- wine developers who wish to subsidise Transgaming's development costs 
> can simply mail their patches to Transgaming, or authorise them to poach 
> them off the wine-patches mailing list (assuming they are not derived 
> from any LGPLed code...).  Ove Kaaven and other Transgaming employees 
> could then apply them directly to the WineX tree.

They could pretty much do that now.  What's to stop them?  At least this 
way there's a chance of the X11 tree thriving.  Maybe that's part of 
Transgaming's goal, beyond the obvious immediate goal of getting code.

> I'm sure that Transgaming's developer resources must be limited, and it 
> seems like a foolish waste of time for Ove and other Transgaming staff to 
> have to maintain two trees, when one of these has no use value -- either 
> to Transgaming, the Wine project, or anyone else.            

The fact that they're bothering at all must mean something, right?

> Everyone must look at the available evidence and draw their own 
> conclusions about Transgaming's ultimate goals and motivations.  For my 
> part, I see Transgaming's relationship to the Wine project (and their 
> customers, in some cases) as being manipulative and exploitive, and only 
> becoming more so over time.

I don't know.  From the evidence I've seen, it seems like they may have 
been honestly trying to make an open-source model work, and found it harder 
in practice than in theory.  It wouldn't be the first open-source model to 
be difficult to successfully implement.

> The above examples of Transgaming's behavior, as well as their 
> unwillingness to share generally useful improvements (DCOM, 
> InstallShield, etc.) with the wine project (outside their core
> competence in the area of game support, which is to be expected) lead me 
> to believe that they regard wine as a *competitor*, rather than a 
> collaborating partner.  If this is true,then the benefits of any ongoing 
> relationship with Transgaming will continue to be one-sided.

It's hard to know, unless and until their subscriber base reaches critical 
mass.  If it does, maybe they'll be the best friends of the Wine project.  
As long as they fail to get close, they'll continue to be viewed with angst 
and suspicion.  Perhaps that's inevitable.

It doesn't exactly encourage the next company to try this open-source 
model, does it?

> I recommend that unless Transgaming is willing to publicly commit to 
> _using_ and making meaningful contributions to the LGPLed codebase, that 
> wine developers should cut them loose (ie: not dual-license patches to 
> wine).  They can then develop their proprietary fork independently, and 
> wine developers will be able to make rational, informed decisions about 
> what areas of wine to contribute to, knowing that nothing will ever be 
> coming back from Transgaming.

Well, this is based on the assumption that they'll never follow through on 
their promise.  Will they, or won't they?  That's a key question.  However, 
it's also a moot point if they never reach critical mass.

If they never reach critical mass, and assuming they don't find another 
business model which is successful for them, I imagine their last act 
before the business goes under would probably be to release all the code 
under the X11 license, if there's no chance of their business surviving...

Why is everyone so quick to assume bad faith without compelling evidence?

Deven