Fwd: Bug#292078: kdebase: startkde does not check lnusertemp's result?

[David Faure]

On Monday 05 February 2007, Juergen Pfennig wrote:
> On Mon 15.01.2007 23:57, Sune Vuorela wrote:
> > A debian user reported this a long time ago to the debian bts, but the
> > issue still applies. (original link: http://bugs.debian.org/293078 - but
> > relevant info pasted below )
> >
> > It has a patch for a issue about not handling possible errors in the calls
> > to lnusertmp in the startkde-script.
> > He writes:
> ...
> > /Sune
> 
> Hi Sune
> just saw your patch but I don't like it. (I guess your patch is intended for 
> kde 3.x). It increases the number of processes that are to be started until 
> KDE is up and running. Even on a stream-lined Debian Etch system around 1800 
> processes get started from boot until KDE is ready.
> 
> So let's try no to increase the number of processes.
> 
> In this case: your patch uses 'which lnusertemp' thereby starting an 
> additional process. Instead it could be more efficient to analyse the return 
> code. Shouldn't bash/dash return 127 if a program was not found? And the [ 
> (or test) operator is a bash built-in and should be a built-in also for 
> dash - dash is Debian's/Ubuntu's incarnation of /bin/sh.
I agree with both points.

> Does the author of startkde read this? 
Funny question; startkde is a collective result from many people's hacks ;)

> Does it make sense to clean these things up for kde 3.x or are these clean ups 
> even relevant for kde 4?
startkde is pretty similar for those things between both versions, I think.
Can you make a cleanup patch for both versions?

-- 
David Faure, faure at kde.org, sponsored by Trolltech to work on KDE,
Konqueror (http://www.konqueror.org), and KOffice (http://www.koffice.org).