Bug#541207: gt.m: changing back from ITP to RFP
On 20/02/11 at 11:42 -0500, Bhaskar, K.S wrote:
>
>
> On 02/20/2011 11:33 AM, Lucas Nussbaum wrote:
> >
> > On 20/02/11 at 10:23 -0500, Bhaskar, K.S wrote:
> > > I manage the group that develops FIS GT.M upstream. Although there has
> > > been no evidence of activity on the Debian lists in the last six months,
> > > there has been activity upstream.
> > >
> > > Here is the issue: just as it takes a gcc binary to compile gcc, building
> > > GT.M has a bootstrapping step and it takes a GT.M binary to build GT.M.
> > > This appears to be a catch-22 situation for GT.M (although it is
> > > apparently not a catch-22 situation for gcc). We have been trying to use
> > > Python for the bootstrap step, but still do not have a successful GT.M
> > > build yet. If you can advise us on a way to break out of the catch-22
> > > situation - what it would essentially take is an initial exception to
> > > allow is to use a GT.M binary to build GT.M from source - we can make a
> > > lot more progress, and progress visible to you. Otherwise, please leave
> > > GT.M in the ITP stage, and once we are able to use Python successfully to
> > > build GT.M, there will be progress that is visible to you.
> >
> > I don't think that this is an issue at all. You could just build GT.M
> > manually on every architecture to do the initial upload. Once there's a
> > version in the archive, then new versions will be built with the
> > previous version.
> >
> [KSB] Are you sure about it. That was not my understanding of how the
> system worked, but if you are right, it does simplify things. Can you
> say definitively, point me to a discussion and/or connect me with someone
> who can? Thank you very much.
Ask debian-devel@ if you want to be sure.
- Lucas
Reply to: