BSD, Gav, LGPL, Jeremy, and business

[Patrik Stridvall]

> First, and formost, we have now heard the oppinion of two big 
> (from Wine
> POV) commercial players: Gav and Jeremy. Now, it was 
> painfully obvious to
> me that Jeremy had the power of reason on his side, while Gav only a
> strong emotional impulse. 

Since choosing the LGPL is a one way street,
it is hard to not be emotional,
when arguing against it.

It is much easier for the LGPL propenents to be emotionally detached.
If you lose now, ou can come back next month, next year or whenever
and try again. The LGPL oppents can't for obvious reasons.

> And there's reason for that -- the 
> LGPL make a
> lot more sense for Wine.

Still think that after reading my reply to Jeremy?

> We've heard countless arguments why that's not that case. A 
> lot of them
> where simply nonsense. The few that made some sense, were ALL 
> based on a
> false premise: that Wine is a monolithic product.

I have never believed that Wine is a monolithics product.
 
> Now, this is a _fundamental_ point that is ignored over and over
> again: Wine is a _collection_ of products, just like a Linux
> distribution. Which means that the viral aspect of the LGPL _stops_ at
> every DLL boundary. 

Very true and this significantly weakens any "protection"
that the LGPL offers.

> It is easy to see now that this does not 
> preclude most
> commercial implementations. In fact, it encourages them givin 
> them a level
> playing field. Just like Linux does.

Perhaps, but at a terrible price. Read my reply to Jeremy.
 
> People, for crying out loud, _think_ about it, and the conclusion will
> jump right out at you. THIS IS WHAT WE WANT:
>   -- if a company invests a _little_ to improve a DLL, we 
> should have no
> moral problem requireing them to contribute that back. Since 
> it's just a
> little contribution, no business will be destroyed because of it.
>   -- if a company makes a huge improvement (like 
> Transgaming), they can
> simply drop the original Wine code, and keep everything 
> propriatary. Now,
> compared to the _huge_ improvement, the cost of 
> reimplementing the mostly
> non-working Wine code should be trivial.
>
> Bottom line is:
>   -- are we raising the bar for comercial companies with a LGPL
> licence? YES!
>   -- how _much_ are we raising it? BY A TRIVIALLY SMALL AMOUNT!
>   -- do we invalidate TG business model? NO!

Somehow you seem to believe that all reasonable business models
will strictly following DLL boundaries, but you have given.
absolutely no proof of this, you just assume it.

Face it the design of the Windows API are a terrible mess that
doesn't follow DLL boundaries.