Re: Should Debian use lsb init-functions?
On Wed, 04 May 2005, Wouter Verhelst wrote:
> > I think it would be better if we simply made rc capture initscripts'
> > standard output (and exit status) and formatted it in such a way that
> > bootup messages were prettier.
> That sounds like an ugly and error-prone hack to me. Not something we
> want one of our most important systems to be working with.
Ugly hack? Not at all, it is good design.
At wrost, you get human-parseable output that, while it does not make enough
sense for the system to color it green or red, won't break anything (i.e.
exactly what we have right now).
All messages from every initscript will be easy to group together (hint:
tagging the origin of the messages is not done by the initscript, but by the
initscript system) and to track, lint, log, etc.
It does not force the use of posix shell for the initscripts, either.
Believe it or not, usage of posix shell for initscripts is not mandated
anywhere, and it would be poor design to force it for no good reason.
As for error prone, have you ever tried to audit our initscripts? A
controlled rewrite of all of them would benefit us a lot anyway. I don't
think echo'ing "ERROR:" or "OK:" to stdout/stderr is any more error-prone
than calling shell functions (which we will provide in the lsb package
anyway, so if someone wants to pre-depends on lsb...)
"One disk to rule them all, One disk to find them. One disk to bring
them all and in the darkness grind them. In the Land of Redmond
where the shadows lie." -- The Silicon Valley Tarot