Interesting article on boot ordering

[Nicolas Mailhot]

Le mer 24/09/2003 à 00:14, Bill Nottingham a écrit :
> Nicolas Mailhot (Nicolas.Mailhot at laPoste.net) said: 
> > - someone plugs the big hole wrt starting daemons as non-root
> > - logs messages are formated sanely (when one compares RH functions
> > behaviour to the ones in LSB with the same intent one can only wonder if
> > the LSB->RH mapping was intentionally botched)
> 
> Well, no one has filed bugs, so if it's botched...

I didn't think it really;) More like "rushed implementation to stick the
LSB logo on the retail boxes that was never improved on"

> Seriously, can you elaborate a little more on these two?

The first one is a big LSB oversight. There is no specified parameter to
the LSB daemon function to specify a specific user (like in the RH
function) and this hurts big time in the real world. (sure one can do a
su separately but why use the function at all if you start rewriting
parts of it ?) This could be taken care of in the next LSB spec revision
if vendors were serious about LSB init scripts part.

The second one is real silly - take any init script and try to convert
all the messages to the lsb functions (don't remember the actual names).
then run it using redhat-lsb. You'll run in no end of problems because
they do a linebreak at the end (I think - didn't test them lately) so
one can not construct a single line from several call. Plus all messages
are attributed in syslog to lsb not the originating service.

The first one is a showstopper. The second problem is only annoying and
showing a lack of polish in the implementation. Many people would gladly
trade a little formating beauty to portability if they could.

(anyway most people try redhat-lsb, see the huge dep list and run).

Current messages implementation problems aside (which are 100% RedHat
fault if my reading of the spec is correct) the LSB service format
itself is not half-bad and could serve as a base if only the daemon as
non-root part could be addressed. 

(if only to remove the stupid checks for the functions file at the start
of each service - with lsb one knows at least the master file exists and
is in a fixed location)

Cheers,

-- 
Nicolas Mailhot
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Ceci est une partie de message
	=?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=
Url : http://lists.fedoraproject.org/pipermail/devel/attachments/20030924/fa4d1123/attachment-0002.bin