Re: Bug#22197: apt: tries to install bin-i386 on m68k, requesting , removal of deb
Jason Gunthorpe <jgg@gpu.srv.ualberta.ca> writes:
> On 8 May 1998, James Troup wrote:
>
> > > #2 Your bug output shows that is is a version eariler than .6 - it links
> > > against slang and other packages that .8 does not, bu the
> > > version says .8! How can this be?
> >
> > Read the bug report, he ran `bug' on an i386 machine, not an m68k
> > one.
>
> That still doesn't explain it, bug showed dependencies on libraries
> that no version of .8 ever had, no matter what machine he ran it on
> those should not have shown.
I give up then. But in any event, it's not a problem with the m68k
deb as it has exactly the same Depends as i386's 0.0.9 (you can check
that for yourself, dpkg -I works on any architecture's debs). If the
issue still bothers you, I suggest you take it up with Goswin. (Maybe
the i386 machine he used to file the bug had an older version of apt?)
[...]
> Well, that's exactly why it should undergo some testing on new
> arches. It is in experimental so it can get some testing from people
> who actually do want to test it. If you are going to port something
> from experimental I think it is more important to give it a test run
> before releasing because that is why it is in experimental - so it
> can undergo testing from people who do want to test it. If you are
> not one of these people then why compile it?
So others who can't compile it, *can* test it. Not everyone has the
space for a compiler&friends or a machine which can easily do
compilation, especially on m68k.
By making the deb available Stefan ensured that the m68k users who
don't/won't/can't compile debs can test it, as you seem so eager to
happen. If it's broken, a bug gets filed. No big deal; right? It's
not as if we're not talking about dpkg, gzip, libc6 or some other
Essential package here.
--
James
--
To UNSUBSCRIBE, email to debian-68k-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: