On Thu, Dec 13, 2012 at 1:53 PM, NightStrike <nightstr...@gmail.com> wrote: > On Thu, Dec 13, 2012 at 2:21 AM, John Marino <gnu...@marino.st> wrote: >> Which clause are you invoking to remove it from the primary tier list? >> Richard claimed "they are not at all happy with GPLv3". That's not a reason >> listed on your reference. He also claimed they "not using still maintained >> compilers" which is also clearly false. >> >> Maybe FreeBSD violates the "Primary platforms are popular systems" clause >> (of course "popular" is not quantified so no way to defend accusations of >> being not popular enough). GCC is certainly "frequently used" on FreeBSD so >> that's not in violation. > > Probably the part about running regular testsuites and posting them to > the mailing list, tracking the regressions, and fixing them.
Gerald runs regression tests on both i386 and x86_64 freebsd (though some old versions of it). We do have a listed maintainer for freebsd. Apart from build issues I am not aware of frequent freebsd specific bugs. If we deprecate i386 it shouldn't stay as i386-freebsd though. There isn't a rule what target can become primary or secondary (but at least it should have a listed active maintainer and regular testresults appearing on gcc-testresults). Recently we dropped ia64-linux for the fact it was actively unmaintained. I'm not arguing in any way against i386-freebsd but if we have to change it I did want to drop the occasional reason against having it on the list (eh, release managers always are happy dropping targets from the list of those they have to block releases for ;)) >From a users view you could argue that singling out freebsd from the other bsds is unfair (freebsd is primary but the others are not even secondary?). There are a total of 6 x86 variant targets on the primary/secondary list (out of a total 13). Maybe we should start to use wildcards there and separate host architectures/OSs we care for from target CPUs and target OSs. And drop the distinction between primary and secondary. (now I'm getting off-topic) Richard.
