[colug-432] CUPS Server

Rob Funk rfunk at funknet.net
Tue May 29 17:39:39 EDT 2012


On Tuesday, May 29, 2012 05:06:21 PM Rick Troth wrote:
> To prolong the hijacking of this thread, I feel the same as Rick and
> Rob that 'service' is annoying.  I would go further and say that it is
> "change for change sake", which is contra-goodness.  I would not
> object if it were simply a wrapper around the traditional INIT
> scripts.  But it seems it is more ... or wants to be.

It started on Red Hat as simply a wrapper around the init scripts, but on 
modern Ubuntu systems it connects to the Upstart init system if there's an 
Upstart configuration for the specified service. (It's a shell script, so you 
can see exactly what it's doing.) Presumably on the latest Fedora it similarly 
controls the new systemd init system, and there seems to be momentum to put 
systemd just about everywhere except Ubuntu. Neither of these newer init 
systems use /etc/init.d scripts natively, and we're just lucky that the 
backward-compatibility is still there for most services.

(There have been lots of big debates on LWN and elsewhere about these 
transitions, particularly systemd.)

>From that perspective, I understand the reasons for using "service" - it makes 
it easier to transition to different init systems that may not necessarily use 
/etc/init.d (or just put the init scripts in a different place, like when RH 
had them under /etc/rc.d/init.d).

BTW, I just noticed that on my Ubuntu system I do have logical tab-completion 
for the service command's first argument in bash. Well, for services in 
/etc/init.d I have it; I don't have it for those that have native Upstart 
configurations but not /etc/init.d scripts. (That's set in 
/etc/bash_completion.d/service on my Ubuntu 11.04 system.)



More information about the colug-432 mailing list