Coin, Steve Langasek <[EMAIL PROTECTED]> writes:
>> I don't see any reason to work around plenty of packages when this is >> for sure an important toolchain bug which cannot be delayed for long and >> my packages are not important ones and are already available >> (unaffected) for most architectures. > > I don't see how being unable to use -O3 is "important". As policy > clearly points out, -O3 is not appropriate as a general-purpose > optimization flag; there are very few packages that actually need it, > and I'm pretty sure it's inappropriate to use as a default optimization > flag for python modules. Not important for Debian, but still a bad brokeness. Any proof there is null probability this bug affects other parts of code ? > Ok, then it sounds to me like the appropriate thing to do is to reassign > this bug to python, rather than to "wontfix" it. Perfect. > I think that fixing a bug that only affects unnecessary optimization > flags should be considered a lower priority than fixing bugs that cause > failures *without* optimization. Don't you? Of course... -- Marc Dequènes (Duck)
pgpqXvz3pd1Dk.pgp
Description: PGP signature
