[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#345973: New cups configuration splitout breaks Kprinter

On Wednesday 04 January 2006 11:27, you wrote:
> Christopher Martin <chrsmrtn@debian.org> writes:
> > Package: cupsys
> > Version: 1.1.23-13
> > Severity: important
> >
> > I'm filing my mailing list message as a bug, since it didn't receive a
> > response when sent to the mailing list directly.
> >
> > In summary, the splitting out of the Port and Listen settings has
> > broken Kprinter. Details below.
> This looks like a serious bug in kprinter: it doesn't know how to
> correctly parse the CUPS configuration file.  While the CUPS split
> configuration shows up the bug, it is not a bug in CUPS, in my
> opinion.
> This should probably be reassigned to kprinter.  If kprinter is now
> actively dangerous, the severity can also be raised.

Kprinter basically reads in the existing settings, then completely rewrites 
cupsd.conf with any user-specified changes. Already a dodgy way of doing 
things, granted, but not trivial to patch.

Given this, consider the complexity involved in it following the "Include" 
files in cupsd.conf. It would have to remember that certain settings 
belonged to each include, and ensure that when written out, these settings 
were written to the relevant include, not the main cupsd.conf. Now imagine 
that a user has commented out a setting in one of the includes. Kprinter, 
when writing its settings, places the setting (which as far as it knows 
wasn't set anywhere) in cupsd.conf. The user now uncomments the settings in 
one of the included files. Which setting counts? Does the user realize what 
they've done?

I understand that you'd rather reassign this bug and tell KDE to re-write 
their printer tools, but I'm not sure that this will yield results. This 
bug is an attempt to at least make the CUPS team aware of what's happened, 
i.e. the full implications of the recent changes, and ask for suggestions 
as much as anything else; if you can think of a solution, then great.

Christopher Martin

Reply to: