Houston, we have a problem.

[Ingo Klöcker]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thursday 05 September 2002 22:59, Carsten Pfeiffer wrote:
> On Thursday 05 September 2002 22:37, Waldo Bastian wrote:
> > > While gpg is rightfully doing it's calculations it has to fight
> > > for CPU time with kmail which is in a busy-loop.  "kill -stop" on
> > > the kmail process allows gpg more CPU time to finish the job at
> > > which time I can do "kill -cont" on kmail to let it run again.
>
> Also if gpg is trying to download a key from a keyserver and the
> keyserver doesn't respond (or takes some time), kmail takes all CPU.
> In the case of the server not responding, forever.

Proposal: Use a more reliable keyserver.

> Being stuck in a select() seems slightly better than polling, to meI
> thought an async solution is being worked on for 3.2, so why not
> commit the select() solution now?

JFYI, the "select() solution" is in fact no select() solution. I just 
thought that this patch used select(), but unfortunately it doesn't.

I will see if I can beautify my ugly code by using KProcess this 
weekend.

BTW, by the time I wrote "This is a known problem. I will fix this for 
KDE 3.2 (it's too late for KDE 3.1)" I didn't know that the feature 
freeze would be postponed for two weeks.

Regards,
Ingo

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE9d9cHGnR+RTDgudgRAutzAKDhHJAxHpL01ADn/0QGjGLCsGnTDACfdMra
ykCsRfA4kcoMJSU0jhctsL0=
=V8a4
-----END PGP SIGNATURE-----