Re: WineX and the AFPL

On 29 May 2002, Alexandre Julliard wrote:

> I'm perfectly willing to collaborate with Transgaming, and any other
> company out there, under *reasonable* terms. I happen to think the
> LGPL is an example of perfectly reasonable terms, and all the other
> Wine companies I talked to (CodeWeavers, Lindows, Macadamian, OsoComm)
> agree with that.

I'm an outsider looking in here, and I would also agree that the LGPL is an 
example of perfectly reasonable terms.  But, having decided to use the LGPL 
now, does that make alternative terms inherently any less reasonable?

> I think it's pretty clear that the obstacle to a collaboration is not
> the license: Transgaming had more than a year before the license
> change to show a willingness to collaborate, yet they didn't do
> anything except making more vague promises, even when it was getting
> clear that many developers were very annoyed with their behavior. They
> obviously didn't feel any need to release significant amounts of code
> as long as Wine code was steadily flowing into their tree.

Yup, it sure sounds like Transgaming wasn't motivated by community spirit.
Then again, most business ventures aren't -- as a rule, they're in it to 
make money, and many will tend not to "give back" much (if at all) unless 
forced to, as with the GPL and LGPL.

> If Transgaming is ready to change to a more open source model, I'm
> definitely willing to go out of my way to help them. As long as their
> model is "we take all we can and release as little as we can get away
> with", I have no interest in continuing such a one-sided "collaboration".

It sounds like the old "collaboration" (before the license change) was so 
one-sided that it annoyed many developers.  Hence, the change to using the 
LGPL license.  It appears to have been successful -- Transgaming may not be 
giving back to the LGPL codebase any more than before, but at least they're 
not receiving a stream of free code flowing into their tree.  0% flowing in
either direction may not be as ideal as 100%, but at least it's equitable.

Since the switch to LGPL, Transgaming has decided it's to their advantage 
(from a business perspective) to offer to release certain portions (code X) 
of their currently-proprietary codebase in trade for the release of other 
portions (code Y) of the now-LGPL codebase, rather than expend the effort 
of re-implementing the functionality of code Y.  That is, they consider the
advantage of keeping code X proprietary less valuable than having (or 
avoiding the cost of reimplementing) code Y.

Now, supposing the value of code X and code Y are roughly equal in value, 
it seems clear that such a trade would be equitable.  As an outsider (and 
hopefully objective thereby), an _equitable_ trade of code seems like just 
as reasonable terms for collaboration as the LGPL is.  Of course, if code X 
is substantially less valuable than code Y, you're back to the one-sided 
"collaboration" that annoyed the Wine developers originally.  On the other 
hand, if code X is substantially MORE valuable than code Y, it could be 
viewed as Transgaming finally "giving back" for taking advantage of the 
generosity of Wine developers under the original X11 license.

Quid pro quo isn't idealistic or utopian, but it's realistic.  And even if 
it's a bit distasteful, it still leaves both sides better off than before, 
and isn't that really the point of using the LGPL in the first place?

> And yes, strange as it may seem, having fun is more important to me
> than Transgaming's bottom line.

Transgaming's bottom line matters to Transgaming, of course.  There's no 
reason it should matter to you.  Similarly, their business model (whether 
good or bad) should also be irrelevant here.  It seems that your concern 
ought to be with the Wine project itself, which is now under the LGPL, and 
with your experience as a developer.

I agree, you should be having fun and writing code, not worrying about 
Transgaming's bottom line.  My question is this: is it really as fun to 
reimplement code that's already been written before as it is to implement 
something for the first time?  As a programmer myself, I have to wonder.

In another message [Re: Proposal for Trade: COM/DCOM for DLL Seperation], 
on May 11, you wrote:

> As I explained already, I think the idea of trading patches is
> fundamentally flawed. I also think Wine will be much better off in the
> long run by sticking to the LGPL than by giving it up now for a small
> short-term gain. So no, I'm not interested in trading any of my
> patches.

I find this stance a bit bewildering, actually.  I fully understand that 
the Wine developers don't want to feel taken advantage of, and there seems 
to be some sort of history with Transgaming where some developers have felt
that way in the past.  Hence, the switch to the LGPL license makes sense, 
and protects you from being taken advantage of.  So far, so good.

However, to reject (out of hand) the very notion of trading patches?  This 
doesn't make sense to me at all.  Certainly, it might make sense to reject 
a proposed trade if you don't feel like they're offering enough -- that's 
negotiation, and they can always come back with a better offer.  But to 
walk away from the table and refuse to even consider cooperating?  That's a 
pretty extremist stance.

While you can get the short-term gain by trading patches, I don't see how 
that translates into "giving up" on the LGPL or its ideals of sharing.  The 
Wine project would remain under the LGPL, just improved from the state it 
would be in without trading.  Transgaming certainly doesn't seem inclined 
to give up on maintaining the X11 fork and switch to the LGPL codebase, so 
what is the LGPL codebase possibly losing by allowing equitable trades?

Must it be "my way or the highway"?  Any given trade can be evaluated on 
its merits.  You can even demand that you'll only take trades where the 
balance of the trade is heavily in your favor (in recognition of the amount 
of work you've already put into the X11-licensed codebase).  What's really 
to lose here?

The LGPL is based on the idea that we should all share our work equally.  
Now, the way it's written, "equally" in this case means that each of us 
contribute 100% of our code back to the community.  Now, this is a great 
thing, but some people aren't willing to go that far.  Evidently, this is 
such a case.  Transgaming evidently doesn't want to contribute back 100% of 
what they develop.  That's their perogative, whether or not it's necessary.

But "equally" could also mean that you give as much as you take.  The LGPL 
doesn't work this way, but it's in the same spirit of sharing and sharing 
alike.  In fact, there's even an argument to be made that this sort of quid 
pro quo is actually more fair all around -- after all, GPL/LGPL software 
(and all other Free Software and Open Source Software) tends to suffer from
the "free rider" problem, especially end-user applications.

At least Transgaming is (finally) offering to make a tangible contribution 
back to the Wine project, in exchange for the value they've heretofore been 
able to get for free.  Is this "too little, too late"?  Your May 11 message 
seems to indicate that you think so.  Maybe this is an emotional response 
to "bad blood" from prior history.  I don't know; I wasn't involved.

As an outsider, it seems to me like you're willing to cut off your nose to 
spite your face.  I don't see the logic behind refusing the idea outright.  
It can only lead to duplication and fragmentation of effort.  Refusing to 
cooperate harms everyone, and isn't that the "evil" of proprietary software 
that the GPL was created to work against in the first place?

Negotiating to trade patches on a quid pro quo basis may not be as ideal as 
automatically sharing everything (as the GPL and LGPL otherwise require), 
but half a loaf is better than none!  Why not take what you can get, since 
it's clear that Transgaming is unwilling to adopt a completely open-source
business model?  If it's an equitable trade, it helps everyone, and isn't 
that functionally equivalent to the kind of sharing that the GPL and LGPL 
are designed to foster?

And if there's still bad feelings against Transgaming in particular, for 
the sense that Wine developers were being taken advantage of, that's fine.  
Make them pay by demanding that any code trades be heavily in favor of the 
Wine project (i.e. Transgaming contributes code substantially more valuable 
than the LGPL code they get in return), to make up for the past.  Continue 
until everyone agrees that they've made up for the past transgressions.  
Whatever makes the Wine developers feel that the arrangement is fair.

But why walk away entirely?  No matter how one-sided the past arrangement 
may have been, "taking your ball and going home" is only able to keep it 
from continuing; it can't make up for it.  At least quid pro quo trading of 
patches can help to rebalance the scales, especially if the trades are 
weighted in favor of the Wine project...

What's past is past.  Why not look to the future and do what's in the best 
interest of the Wine project?  Quid pro quo trading of patches doesn't 
violate the _spirit_ of sharing enshrined in the LGPL, even if it doesn't 
match the letter of the license.  As long as trades are fair and equitable, 
how could it be a bad thing, or endanger the future health of the project?

Maybe I'm missing something here.  Like I said, I'm an outsider.  I should 
probably have stayed out of it, but it just doesn't make sense from this 
vantage point.  What am I missing?  What's so wrong with trading patches?

Deven

P.S. I'm not stating any opinion as to whether Transgaming's business model 
is good or bad, or whether or not they took actually advantage of other 
Wine developers.  I was only addressing the _perception_ that some of the
developers seem to feel like they were taken advantage of.  Those feelings 
are valid whether or not an objective observer would say that Transgaming 
took advantage.  As for their business model, I haven't evaluated it, but 
I'm not going to presume that it's a bad business model solely on the basis 
that they refuse to release all of their code back to the Wine project.
After all, for a business to pay developers, they have to have revenue from 
some source, and no completely open-source model has proven foolproof...

P.P.S. Sorry for rambling so much!