Re: eggcups (and libgnomecups) for 2.12

[Matthias Clasen <mclasen redhat com>]

On Wed, 2005-07-20 at 11:19 -0600, Andreas J. Guelzow wrote:
> On Wed, 2005-20-07 at 13:16 -0400, Matthias Clasen wrote:
> > > > Why wouldn't one have a local cupsd?
> > > 
> > > Good question. We have local cups daemons. Since the whole asynchronous
> > > stuff apparently got included to avoid delays when talking with a
> > > non-responsive remote cups daemon, I assume there are some people
> > > somewhere that are trying to avoid having a local cups daemon.
> > 
> > No, the async stuff is necessary even for local cupsd, since the local
> > cupsd can get stuck for noticeable amounts of time while talking to 
> > remote servers...
> 
> How? Perhaps I don't understand all the details, but the critical issues
> should be
> 
> a) asking the local cups daemon which printers are available. The local
> cups daemon should be able to respond without talking to a remote
> printer.
> 
> b) submitting a print job to the local daemon. Again it shuold place the
> job in its local queue and only then try to talk to the remote daemon.
> It does the latter asynchronously anyways.
> 
> c) loading ppd files. Yes they need to be retrieved from the remote
> daemon, but this should happen only when the printer is selected in the
> dialog. So in the sychronous way that will cause a delay if the user
> wants to print to a misconfigured or down printer and that printer is
> still known to the local daemon, which in a broadcast situation will
> only happen for a short period of time after the printer went down.


See, what you are describing is the ideal world, with a
non-sucking-cupsd. Working on eggcups was a good way to learn that the
world is not ideal, in particular wrt to cupsd...