gdm

Dagmar d'Surreal dagmar.wants at nospam.com
Wed Jun 18 15:35:13 PDT 2003


On Wed, 2003-06-18 at 04:01, Michael A. Peters wrote:
> On Wed, 2003-06-18 at 01:22, Dagmar d'Surreal wrote:
> > 
> > Init has rudimentary functions for detecting loops, however, its
> > "response" is just to "stop doing that".  This is not what I would
> > consider to be a particularly flexible or useful behaviour, especially
> > if you didn't already start sshd or some hidden agettys so an admin can
> > try to fix things.  There is no downside to starting GDM through a
> > script, dude, _certainly_ no ways to "accidentally" interrupt them
> > unless your system is tragically misconfigured, and plenty of reasons
> > _for_ invoking shell scripts through init.
> 
> starting gdm in the inittab guarantees that all other init scripts are
> run first. sshd (if you run it) would be started before gdm.

This has absolutely nothing to do with what I've been saying... that you
should not start gdm directly through init, but instead call a shell
script that invokes gdm and can handle failures.

> You also should be able to ctrl-alt to a virtual console and login
> through one of those should there be something like a missing library.

Yes, another wonderful architecture decision in opposition to the idea
of unattended and _remote_ reboots.  However, this has very little to do
with why one shouldn't invoke gdm directly from inittab.

> by adding it the end of the inittab (as I suggested) your inittab would
> look like this (at the end):
> 
> 1:2345:respawn:/sbin/agetty tty1 9600
> 2:2345:respawn:/sbin/agetty tty2 9600
> 3:2345:respawn:/sbin/agetty tty3 9600
> 4:2345:respawn:/sbin/agetty tty4 9600
> 5:2345:respawn:/sbin/agetty tty5 9600
> 6:2345:respawn:/sbin/agetty tty6 9600
> 
> # Run xdm in runlevel 5
> x:5:respawn:/usr/bin/gdm
> 
> As you can see - /sbin/agetty is started on tty[1..6]
> This happens _before_ gdm is spawned.

Man, all these years of working with fistfuls of unix machines and I
never noticed they executed commands in sequence from top to bottom.  If
you think this sarcasm is unwarranted, consider that you're talking down
to someone with over 15 years experience banging on Unix machines.

> This why you will see a "console" login _before_ gdm is spawned.
> The only issue is the case where there is no mouse so gdm (after taking
> over on tty7) decided you don't need a keyboard either.

...which has nothing to do with why one shouldn't start gdm directly
from inittab, AND is entirely insane.  GDM doesn't give a tinker's damn
about your mouse or your keyboard, that's XFree86's job, as I said in
the last email.

> This would also happen in the case of the starting gdm from an init
> script.

...which has nothing to do with why one shouldn't start gdm directly
from initab.

> I don't mean to be rude, but your objections to do it from the inittab
> don't have merit.

I don't mean to be rude, but you don't seem to be paying attention to
what I've been typing anyway.  You're dead wrong.

> If there is a messed up XFree86 config file, gdm will be graceful in how
> it handles it, even offering to run a tool to configure it.

This is one of the things *I* have been saying, but clearly you've not
been paying enough attention to notice.  However, and this is just a
theoretical possibility here, and I know it's really going out on a limb
that you'll even hear this, but (big letters for high visibility, not
shouting) WHAT IF GDM ITSELF IS BROKEN IN A MATTER ENTIRELY UNRELATED TO
THE XFREE86 CONFIGURATION...

> If the mouse isn't attached, there is no difference - currently you are
> screwed either way. gdm will still start and switch you to tty7 and
> leave you without keyboard capability.

...which has nothing to do with why one shouldn't start gdm directly
from inittab.

> If you remove a library, agetty is still spawned on tty[1..6] will
> already have been spawned, and you can log into your system through one
> of those consoles.

...which has nothing to do with why one shouldn't start gdm directly
from inittab.

> What blfs decides to do - I could care less.
> But there is an easier way to boot into gdm than the method currently in
> the blfs book. And there isn't a reason not to do it the easier way.
> Other than "we've never done it that way before"
> 
> I'll just leave it at that.

...which has nothing to do with why one shouldn't start gdm directly
from inittab.

I'm going to try one last time...

x:5:respawn:/usr/bin/gdm  <-- BAD.

x:5:respawn:/etc/init.d/rc.5  <-- GOOD.

If you STILL don't get it, don't bother responding. because I sure as
hell won't again.  This is absolutely ridiculous.

-- 
The email address above is just as phony as it looks, and for obvious reasons.
Instant messaging contact nfo: AIM: evilDagmar  Jabber: evilDagmar at jabber.org

-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe blfs-dev' in the subject header of the message



More information about the blfs-dev mailing list