Re: Alternatives to gcc (was Re: gcc 4.3: when will it become standard compiler?)

Michel Talon wrote:
> Garrett Cooper wrote:
> 
>> On Sun, Jan 11, 2009 at 4:31 PM, Eitan Adler <eitanadlerlist_at_gmail.com>
>> wrote:
>>> I never took care about GPLv2 and v3
>>>> differences but know, this seems to come to relevance in some way.
>>> I don't seem to understand this.  Why should gpl v3 affect the OS?
>>> The
>>> output of the compiler isn't affected by the license.  Is it?
>> Yes the GPLv3 is `extremely viral' when dealing with proprietary
>> innovations and features, compared to GPLv2.
> 
> This is complete rubbish, the V3 is viral in exactly the same way as the
> V2, the V3 simply closes a hole allowing people to do things which were
> obviously against the spirit of the V2, but were able to go through a
> hole in the letter of the V2. The spirit of the GPL has always been: "if
> you derive software from GPL software, and you distribute it, then it
> must be distributed under the GPL, and you must provide source code.
> The end users can modify the code, distribute their work, and run the
> modified version as they see fit". The Tivo loophole was that the
> hardaware runs only signed binaries, hence you cannot run your modified
> version. This is completely obviously in contradiction with the spirit
> of the GPL, if allowed by the letter. Now the large incertitude in this
> game is to know what is exactly a derived software from another one.
> This is an extremely vague concept.
> 
>> Many companies also have to write in functionality and tie-ins which
>> expose portions of the OS, debugging tools, libraries, etc that would
>> require them to expose their proprietary secrets. They should be just
>> as exposed with GPLv2, but the GPLv3 is more stringent and the FSF is
>> ramping up copyright infringement notices
> 
> Precisely they were just as exposed with the GPL V2, but were
> deliberately ignoring their obligations. the fact that the FSF ramps up
> its legal action is the new factor of importance. These legal actions
> also serve to define and precise the concept of derivation, and perhaps
> commercial companies begin to understand that they are vulnerable.
> 
> Anyways the fact that the compiler is GPL V2 or GPL V3 is totally
> insignificant with respect to FreeBSD, since it is an established fact
> that the result of the compilation is not derived from the compiler,
> hence is not subject to the virality of the licence. I think this
> is not so clear for C++ code, because the compiler then includes some
> stuff of its own in the result, but there is perhaps a special exemption
> for that. The last thing FreeBSD needs is losing time and performance to
> solve a problem which doesn't exist in the first place, when there are
> so many actual and concrete problems which are not fixed.

I am not a lawyer.

But I was reading the GPLv3 license.  And I was coming to exactly the 
same conclusions as Michel.

Also, reading the prior postings where Linux kernel developers were 
complaining about GPLv3.  But their complaints were against draft 
versions, and as far as I can see, their complaints were adequately 
handled in the final version.

As for Michel's point that the results of the compilation are not 
covered by GPL - this seems to be stated explicitly in the GPLv3 license.

Stephen