On Tue, 18 Dec 2001, Patrik Stridvall wrote:
You forget that some "independent"(*) parts like the Crypto API are parts of other DLL:s (like ADVAPI32.DLL) for no particular reason.
This is ridiculous: it is one of the few exceptions, it is simply silly to bring the Crypto API into this discussion. If this the Crypto API is the only problem, we can fix it in multiple ways.
It is not the only problem it is just a symptom of the gaping hole in LGPL.
At this point, I would like to know if people agree up to this point.
I don't.
That's informative. WHAT don't you agree with? In fact, if we can agree on WHAT we disagree, I think it would be a great step forward.
"I don't" primarily refered to the previous argument concerning "independent"(*) parts.
- Isn't Wine's best interest to evolve and develop as fast
as it can?
Yes.
This is a fundamental point. A project like Wine is just like a living, breathing creature. And being need to be selfish to survive. Not too selfish, but it is essential that they have a certain level of selfishness.
Your analogy is flawed, Wine is not living in the meaning that it, if it is to unselfish can die. The mechinisms are entirely different compared to normal life.
In Wine's case, that has to be: 'the licence should be such that it would maximize Wine's development'.
Yes, but I don't believe that LGPL does, that I have being trying to explain in I forgot how many mails.
- If so, isn't the LGPL _spirit_ in Wine's best interest?
No, not nessararily. See different mail.
1 follows from 0. Period. You can not agree with 0 and disagree with 1, no matter what you write in other emails! :) Again, THE SPIRIT of LGPL, not the letter. Again, the spirit is: 'if you improve Wine in minor ways, please contribute back your improvements'.
Yes, and the key is minor ways and I believe that is a absolute non-issue as I have explained before. There is no profit in holding minor improvements.
- If so, why shouldn't we formalize it in the license?
But it isn't nessarily so.
And why is that? Quite the contrary, because:
-- For people that contribute back, would chnage nothing -- For people that don't contribute their changes, it's no good to us, so we don't care (which follows by point 0).
Tertium non datur. That is, there's no other case. QED.
OK, since you bring mathematics in to it, I will try to explain in mathematical terms:
"The set of people is not invariant over a license change."
In fact the set of people are likely to be considerly less. While I have no strict mathematical proof of that, I especially point to Gavriels' reply concerning Corel and Transgaming.