On Fri, Jan 09, 2004 at 12:24:45PM +0900, GOTO Masanori wrote:
> At Sat, 3 Jan 2004 21:32:58 -0700,
> Joel Baker wrote:
> > > I discussed with Miquel van Smoorenburg (initscripts maintainer) that
> > > he would to integrate this mount operation from
> > > /etc/init.d/mountkernfs (in glibc package) to
> > > /etc/rcS.d/S35mountall.sh (in initscripts package) or so.
> > >
> > > I basically agreed his idea because this mount operation in
> > > /etc/init.d/devpts.sh or /etc/init.d/mountkernfs are not glibc related
> > > issue. All mount operations should be integrated into one package. I
> > > think Miquel will mail this issue soon in detail.
> > >
> > > So I think it's good idea to wait releasing newer glibc package to sid
> > > before our discussion reaches an agreement, because my mountkernfs
> > > works are still included in the current debian glibc cvs.
> >
> > You can short-stop a bug that will show up, eventually, if you care to. The
> > S35mountall.sh script should check that it's on an OS that has some concept
> > of devpts, usbfs, usbdevfs, etc, before actually trying to mount them...
> > (or maybe it already does, and I missed it, in which case I'll go hide in a
> > corner over here).
>
> Exactly, I think too. /etc/init.d/devpts.sh checks OS version, but
> mountall.sh in initscripts does not check. Miquel, please check and
> fix it if it's not considered for other OS like bsd* and hurd.
>
> BTW, I wonder why this mountall.sh is ok for bsd* and hurd...
Quite possibly because we haven't gotton far enough that it was obvious
that it was breaking, amidst all the rest of the stuff in the init scripts
that probably can't be trusted without inspection? :)
Argh. Must finish moving stuff off of the old box, and reclaim it for BSD
porting (P4 2.6 w/ 160 gigs of disk, vs. a 2x Celeron 400, w/ 20ish, in
the old box...)
--
Joel Baker <fenton@debian.org> ,''`.
Debian GNU/NetBSD(i386) porter : :' :
`. `'
`-
Attachment:
pgpdFfGdqSpCM.pgp
Description: PGP signature