Re: Bug#15859: libc6 in stable is horribly broken
On 12 Dec 1997, Martin Mitchell wrote:
> Chris Fearnley <cjf@netaxs.com> writes:
>
> > >Huh? The upgrade path is quite clear: install a newer libc5 (5.4.33-7)
> > >from hamm, then you may install libc6.
>
> This is the correct upgrade path, perhaps the howto needs to be clarified
> on this point.
Installing libc5 from hamm forces you to abandon your old libc5
development system since it CONFLICTS (correctly) with libc5-dev. Not
everyone is going that route yet.
> > The reason for my bug is to get the broken package off the ftp site.
> > Before anyone else breaks their system. Guy, if everyone believes that
> > 5.4.33-7 in hamm solves the problem, could you replace
> > libc5_5.4.33-6.deb with libc5_5.4.33-7.deb? I won't be hazarding the
> > upgrade to libc6 until tomorrow so I have no opinion that -7 is any
> > good.
>
> The 5.4.33-6 package is _not_ broken, and should not be removed.
> It rightly conflicts with libc6 due to the different utmp format between
> libc5 and libc6. The 5.4.33-7 package in hamm has modified utmp routines
> so it can coexist with libc6.
Okay there is a different utmp format. Lets try to list the packages from
libc6 that care about utmp and would actually mangle it if running with
the old libc5 utmp format:
login
telnetd (maybe, i think it hands off to login)
sshd
rlogind
last (well, it will show the mangled data)
who/w (ditto)
ftpd
> There is nothing wrong here, except perhaps a lack of clear documentation.
The problem is that there are many people who don't have a problem with
the minor issue of possible utmp corruption (which will only happen if you
install something that is compiled with libc6 and does utmp stuff), but
have a MAJOR PROBLEM being FORCED to ABANDON THEIR OLD DEVELOPMENT
ENVIRONMENT.
--
Scott K. Ellis <storm@gate.net> http://www.gate.net/~storm/
--
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: