Re: [PATCH] libkdeinit symbol pruning
Karl Vogel <[email protected]>
| Newsgroups | gmane.comp.kde.devel.optimize |
|---|---|
| Message-ID | <[email protected]> |
Karl Vogel wrote: > On 19 Jul 2004, you wrote in gmane.comp.kde.devel.optimize: > >> On Monday 19 July 2004 14:16, Waldo Bastian wrote: >>> I don't mean linking, I mean plugins that have unresolved symbols >>> that happen to be provided by the application that loads the plugin. >>> But maybe we got rid of such plugins already indeed. >> >> Ah, and obviously those symbols currently end up being found in the >> kdeinit module, linked to by the binary, so the symbol pruning would >> break it. You're right, this might happen. Well, that's a good reason >> for testing this locally and only applying after 3.3. > > NOTE: yesterday evening I did have a few things that acted strange.. don't > know if it was caused by this patch or some other stuff I did to my KDE.. > and it was already late, so didn't really look for the cause. Will try to > have a look at it this evening. Right.. after strugling with CVS HEAD, which appears to be b0rken at the moment [1] , I checked out KDE_3_3_0_BETA_2 kdelibs/kdebase and rebuild with and without the kdeinit pruning.. And the patch indeed seems to be causing a few problems here and there (so far kcookiejar has probs and the kicker preferences button). [1] if only KDE used tinderbox -- so that the guilty parties can be tracked down and shot.. or at least ridiculed a bit :)