Re: Bug#19691: lprng: include a PATH= statement in the init.d script

[Ian Jackson <ian@chiark.greenend.org.uk>]

Adam P. Harris writes ("Re: Bug#19691: lprng: include a PATH= statement in the init.d script "):
...
> Hmm.  I kinda agree.  I don't think init scripts should mess with PATH.
> Unless they have unique requirements.  I think generally the init scripts 
> should assume that path is set correctly.  If there are problems with the 
> path being set by init during boot time, that would be a bug on init.  If 
> there's problems in a sysadmin not having the right path and running an 
> init script, that's an operator error.

I agree entirely.  Can we make this policy ?

> As far as the invokation of the actual daemon in question, I think the 
> path to the binary or script should be absolute, i.e., /usr/sbin/crond or 
> whatever.

I'm not sure about this.  If the daemon is on the PATH (including
/usr/sbin) then it might be useful to execute it from it.  On the
other hand, start-stop-daemon needs an absolute path.

Ian.


--
To UNSUBSCRIBE, email to debian-policy-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org