[SM-Discuss] init.d defaults are evil and bad
Seth Alan Woolley
seth at positivism.org
Tue Dec 14 22:26:39 EST 2004
On Tue, Dec 14, 2004 at 07:15:17PM -0800, Andrew wrote:
> On Tue, Dec 14, 2004 at 07:05:40PM -0800, Seth Alan Woolley wrote:
> > It can, there's an ESSENTIAL="yes" line that can be added to the init
> > script. It's been in sorcery for a long time (hence everybody forgot
> > about it).
> > But, that means it is essential on EVERY system and cannot be disabled
> > except by prodding (an extra confirmation). Failures in essential init
> > scripts will complain loudly during init process (and might even do
> > more).
> > I wrote the initial logic for the default to no except essential
> > scripts. The system asks all these questions when you start casting
> > stuff. It's best to heed them.
> > And, I'm wondering how it turns the box into unusable? If it does THAT,
> > then it is clear we need either an essential init script added or some
> > more logic added to an already existing essential init script. Thus,
> > bug, file, now, as Andrew says. ;)
> The other thing is, that not all the non-ESSENTIAL init scripts in the
> init.d spell are really optional either. Most systems should have them,
> but then when sorcery looks at the ESSENTIAL bit, it doesnt know how to
> make that distinction. The init.d querying code in sorcery was written
> for the common case of a spell installing some non-essential service,
> not for the specific case of the init.d spell, which obviously isnt
> installing your average everyday plug-in init scripts. So what I would
> prefer is the init.d spell do its own thing for init files that are
> optional only for users who really really dont want them. It can and
> should still use the libinitd calls for installing the files but should
> probably have its own specific set of querying routines for these files,
> otherwise we have to make sorcery even more complicated to handle not
> only common case situations but special case situations.
I presume jkolb can tell us where we goofed on defaulting to no at (in a
special case that requires a special "dispensation").
Seth Alan Woolley [seth at positivism.org], SPAM/UCE is unauthorized
Key id EF10E21A = 36AD 8A92 8499 8439 E6A8 3724 D437 AF5D EF10 E21A
Security Team Leader Source Mage GNU/Linux http://www.sourcemage.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: not available
Url : http://lists.ibiblio.org/pipermail/sm-discuss/attachments/20041214/69bf1b9c/attachment.bin
More information about the SM-Discuss