Re: installing gnupg
On Sat, Mar 03, 2001 at 03:16:42PM +0100, Santiago Vila wrote:
> Marcus Brinkmann wrote:
> > Santiago Vila wrote:
> > > 3.5. Make makedev Arch: any, and modify it so that under hurd-i386 the
> > > package does nothing (i.e. it is really a dummy package). It has an
> > > Installed-Size: of 72Kb. Even if we multiply this by the number of
> > > architectures, I think we can live with that (we already live with an
> > > "Arch: any" base-files, which is not a large package either).
> >
> > This is a kludge which will work for makedev, but that's all. What about
> > linux kernel-source packages, which clutter our archive, and I haven't even
> > attempted to make a full list. What about a hurd-doc package?
>
> IMHO, kernel-source should be renamed to linux-source, and people
> willing to cross-compile Linux from Hurd should have these packages
> available (not that someone is really willing to do that right now,
> but who are we to tell people not to do that?). Same for hurd-doc. A
> user might want to install documentation for hurd-doc on a Linux
> system, why not?
Now, I think you are talking your way out of it :) It is out of question
that the linux kernel source can be useful on the Hurd, and the hurd docs
can be useful under Linux, but at a big scale, a feature to say "this is for
all linuxes, but not hurd" is necessary. I am also not convinced that it is
a good idea to actually make those packages "all" rather than "all-system",
but I am not too worried about it.
I also don't think that dummy packages are the way to go. They are used to
satisfy the packaging system, which is already causing the problem. So eliminating
the problem instead is much preferable. I still have this strange motivation
to solve technical problems with a clear design rather with a set of patches
on an old and broken system.
> > I have a different work around for this planned. I will make the hurd
> > package conflict: makedev (< 9:999), and have a makedev dummy package in the
> > alpha.gnu.org archive. That will work for now, keeping the place of kludging
> > small.
>
> Mmm, what does makedev maintainer think about this particular problem?
He said to me that he prefers to be able to say linux-all rather than
changing it to list all linux arches, although he would do this as a work
around if necessary for us. I didn't ask him, because hiding the problem
temporary is not helping but harming us. Instead, for makedev, I will take
the route to make the dummy package and conflict in the Hurd package as
described. This solves the immediate dependencies problem without hiding it.
Thanks,
Marcus
--
`Rhubarb is no Egyptian god.' Debian http://www.debian.org brinkmd@debian.org
Marcus Brinkmann GNU http://www.gnu.org marcus@gnu.org
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de
Reply to: