Re: Bug Reporting system
Goswin Brederlow <email@example.com> writes:
[ Please don't Cc: me on messages going to debian.devel ]
> > > > > My proposal is to add an architecture field to the
> > > > > maintainer and having one maintainer for each architecture.
> > [...]
> > We have ~6 active (pushing it) maintainers for m68k, other
> > non-i386 architectures have a similar number *or less*, just how
> > many packages do you expect to have "architecture maintainers"
> > for? I say again, it's not workable.
> There was a list with maintainers of m68k stuff lately. Something of
> 20-30 packages.
No. No. No. Your ability, to _not_ read what stuff actually says but
to simply read what you want to see, is astonishing.
That list was over a year old and of the people on that list (of 5
maintainers), only one is at all active at the moment and he accounts
for two (2) of the 68 packages (how did you get 20-30?) listed.
> SO what?
So, you want people to go to the hassle of adapting the bug tracking
system and dpkg-* to support architecture-specific maintainers for
20-30 packages (your flawed estimate), which is what? 2% of the
Packages (real figure == .2%)? It's not worth it.
Bug reports for non-i386 architectures is a problem, and some of the
things suggested (like Cc'ing arch-specific bug reports to an
$(ARCH)-bug or -devel list) are plausible and possibly a good idea.
Architecture-specific maintainers is not a good idea and not even
vaguely feasible. Please, give it up.
> The i386 maintainer wouldn't be bothered by some bugs of other
> architectures (like xpaste m68k depending on X11R6 which is
> obsolete). How could I report a bug like xpaste at the moment?
The way you did, by asking on debian.m68k for someone to recompile it.
(Yes, no one's done it yet, but the fact that one can get round the
bug by using --force-depends, and that there are 575 other packages
that need recompiled and that only one active m68k maintainer is still
libc5 based, probably has something to do with that)
> > > > Oh, yeah and it would also break completely if/when
> > > > auto-compilation gets under way.
> > [ ... ]
> > Read what I wrote. I said it would break *with* auto-compilation,
> > since then there are no architecture maintainers any more.
> > [ ... ]
> If it breaks maintainer fields for architectures it would also break
> the maintainer field as it is now.
I never said it breaks maintainer fields, I said there "are no
architecture maintainers". Again, please read what I write or don't
bother replying to it.
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
Trouble? e-mail to firstname.lastname@example.org .