[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: revised proposed solution (was Re: Bug#15859: libc6 in stable is horribly broken)



On Sat, 13 Dec 1997, Remco Blaakmeer wrote:

> On Sat, 13 Dec 1997, Scott K. Ellis wrote:
> 
> > The problem is that libc5-dev doesn't exist in hamm.  Hamm has
> > libc5-altdev instead.  This forces people who want to compile libc5 stuff
> > into the altgcc/lib*-altdev mode, requiring the mass removal and
> > installation of a whole set of development packages.  I'm against forcing
> > people into that just to install "a couple of packages" from hamm.
> 
> Didn't I reply to your reply to the proposal that there would come a hamm
> libc5-dev? I believe you agreed with that proposal in other messages. And
> though I am not at all an expert on this, I think that this would have
> been the right solution, especially if you had started this discussion
> about six months ago, when hamm just started to be developed:
> 
> bo: libc5
> bo: libc5-dev, depends on bo-libc5, provides and conflicts with libc-dev
> 
> hamm: libc6, conflicts with bo-libc5
> hamm: libc6-dev, depends on hamm-libc6, provides and conflicts with libc-dev
> hamm: libc5, different (higher) version from bo with modifications needed
>       to work together with libc6
> hamm: libc5-dev, depends on hamm-libc5, provides and conflicts with libc-dev
> hamm: libc5-altdev, depends on hamm-libc5, conflicts with bo-libc5-dev and
>       hamm-libc6-dev, provides (probably) libc5-dev
> 
> This would make the upgrade path easier, IMHO: first upgrade libc5 and
> libc5-dev (and ldso, I think), then install libc6 and other packages.

If someone will build this setup, I'll happily withdraw my objections.
This seems to be the best of both worlds.  Although the dependency on
libc6 by libc5 still appears spurious and should probably be removed.


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org . 
Trouble?  e-mail to templin@bucknell.edu .


Reply to: