* Nathaniel Smith said:
> Isn't it easier in the local case to just install sash? Then no matter
Yes, it's a solution, but only when sash is installed as a part of the base
Debian installation. In any case, it should only be used as a single mode
shell. For that I'd suggest one of two solutions, please comment on both of
them:
1. modify sulogin to invoke /bin/sash in addition to all the attempts it
takes to provide the admin with a functional shell.
2. (Easier and more elegant). Every Debian installation uses one or another
bootmanager. Why don't their maintainers add an equivalent of:
(lilo) -b rw init=/bin/init.static sushell=/bin/sash
to their bootmanager standard menu? That would provide the user/admin with
an immediate solution to crashes related to dynamic linking. For that to
work, also init should have a static version in /bin.
> what happens, as long as you have a kernel and a filesystem, just add
> init=/bin/sash to the kernel command line (eg, LILO:), and you have a
Well, it works, of course, but there's a little problem. Many users who have
linux at home don't know much enough about linux bootup process and when the
crash happens they don't have the opportunity to fix it anymore. Since they
didn't know about it before, they didn't make any provisions to protect
themselves from the situation. Also, the standard LILO installation is
non-interactive in that it boots directly the linux kernel from HDD.
[snip]
> Making it work remotely is harder, but all it really needs is a static
> ssh, maybe /sbin/login (doubt it), and root's login shell set to sash.
If the system needs physical attention it's foobar anyway...
marek
Attachment:
pgpvVZrC93RvW.pgp
Description: PGP signature