Re: using bugzilla for libc

[Roland McGrath <roland at redhat dot com>]

> Is it possible to have it also automatically cc: based on
> the host triple selected?

Not so far as I know.  You can always do regular queries for
newly-appeared/changed bugs matching some fields you are interested in.

> It would be useful to have some of us involved with Hurd development able
> to see any bugs filed here automatically.  I imagine other non-Linux
> targets would probably find this useful, too.

I think hurd should be a component, referring to the hurd-specific code.
I forgot to put that on my initial list.

> Should we have a keyword or component for designating a function
> that's also in gnulib?

I don't really know what keywords mean.  The functions copied by gnulib are
a random smattering of unrelated things as far as libc or a user is
concerned.  I don't think this kind of thing is the most useful way to make
sure changes affecting gnulib get tracked.  gnulib maintainers can file
bugs in libc bugzilla for changes they need in the libc files they copy, if
that is useful for their tracking.  For gnulib maintainers to keep abreast
of changes in those files, they should simply be watching ChangeLog or the
cvs commit mail for the files of interest (you could set up something
automatic to do this pretty easily).

> For those who haven't seen it, gcc has a webpage[0] showing
> what they consider the various severities to mean.
> 
> [0] http://gcc.gnu.org/bugs/management.html

Everything on that page is reasonably sensible for glibc as well.
A little of what it says compiler-specific, but it's obvious how to map it.



Thanks,
Roland