Re: UML cooperation

[vio]

* Andrew Hill <address@hidden> [001030 10:47]:
> > Let's put it like this. The gnue project is my first encounter with a GNU 
> > project, so I didn't know how you do things around here. Let's say that 
> > FSF's policy of excluding non-open source tools from your toolset comes as 
> > a big shock to me. This simply because my concept of "freedom" goes 
> > something like "the freedom to use whatever tool I (I would like to 
> > emphasize the "I" here) choose" (including non-open source). While in your 
> > concept the choice is not made by me, but by FSF. Hence my shock: I just 
> > don't appreciate that choice being taken away from me for whatever 
> > well-intentionned reason. I still don't get (probably never will) why FSF 
> > cares "how" I build my code. As long as that code is released under GNU 
> > license, what does it matter ??? (I know Derek's answer, it's just not very 
> > convincing for me).
> 
> The freedom of free software is for the benfit of the user, not the
> developer. 

We seem to be running in circles here, arguing on semantics. Nevertheless, 
allow me a small answer.

First, if the gnue team is comfortable with these constraints, good for you 
(use what you are comfortable with first and foremost !!). 

Second, I believe that ALL developers are also users (with some exceptions, of 
course), especially in the open source realm, hence your argument is only 
semantics and isn't relevant, in my view. 

Third, I still can't see the legal relationship FSF is implying between "code" 
and "tool" used in the development of that code. You nor Derek are making the 
case for why (perhaps because this is getting out of scope for this list, and I 
must agree), if I want to release my source code under GNU license, I MUST 
exclude all non-GNU development tools. Judging a tool NOT on technical merits 
but on whether it releases its source code public is plain ridiculous.

What prevents me from releasing MY C++ code under GNU license even if I 
developed with tools like RR or Microsoft's VisualC++ ? It's My code (not 
Microsoft's or Rational's, they just sell tools !!!). Even if I had to pay for 
the tools I used in developing it, I own copyright over it (not MS or 
Rational), so I alone can decide to release it under the license of my choice. 
I really don't see why FSF believe this code has become tainted and is less 
worthy (I'm not judging here the technical merits) than code developed with GNU 
tools. 

In other words, I continue to believe that the legal relationship you are 
implying between licenses on source code and the licenses on the tools used in 
developing that source code is non-existant, and neither Microsoft, Rational, 
Sun (for its java tools), or whoever, is making any legal claim whatsoever over 
code developed with their tools. If they attempted such a scheme, they'd be out 
of the development tools business in no time. Again, apologies to those who 
find this post out-of place on this list. I know gnue won't change its ways of 
doing things, and you shouldn't if you're comfortable with it.
 
A better explanation to this situation may come from a more strategic theater: 
FSF is apparently enforcing a policy of discriminating against non-GNU code 
amongst their ranks, in a move to favor the development of open source code. 
This would make more sense to me. Good. But since I am not on FSF's payroll 
(figure of speech), I am free to express my opinion that technical merit should 
outweight licensing policies, and not the other way around (or FSF's way). This 
being said, go gnue !

Regards - Vio