On Thu, May 24, 2012 at 06:08:16PM +0100, Adam D. Barratt wrote: > On 19.05.2012 19:04, Adam D. Barratt wrote: > >I'm not sure we've ever released with an architecture which was in > >either broken or fucked, but hopefully someone will correct me if I'm > >mistaken on that. > > Anyone? :-) > > Opinions as to whether it makes sense to release an architecture in > either of those states would also be welcome. > I do not think it is sensible to release an architecture that is in broken/fucked. That's what something like debian ports is for. In order to release hurd, even as a tech preview, we need hurd in testing and users actually testing it. This is a problem at this stage because: * there isn't a functional D-I port yet * it doesn't support debian style networking (ifupdown etc) * it doesn't support any meaningful available new hardware (USB, SATA) * its archive coverage is far lower than required Thus, I do not see how we can release with the architecture. More precisely, I do not think that the architecture will give our users the same support and stability as any other architecture in the stable release, and I think that the architecture's inclusion will negatively impact the release process as a whole. Hence, I have updated the architecture release table (http://release.debian.org/wheezy/arch_qualify.html) to mark hurd as 'no' as a candidate for a release. I'm aware that this will not be the news that is wanted, but I do believe that it is the correct decision, and it would not be right to delay this further. Neil
Attachment:
signature.asc
Description: Digital signature