Removing parts from xfce4gui4

Brian J. Tarricone bjt23 at cornell.edu
Mon Nov 19 00:20:23 CET 2007


On Sun, 18 Nov 2007 21:47:42 +0100 Nick Schermer wrote:

> 2007/11/18, Brian J. Tarricone <bjt23 at cornell.edu>:
> > Right, but apparently it can be worked around -- Stafan's tray
> > proves that.  We have a long history of working around various bugs
> > in our systray, and it's somewhat sad that we've lost all those
> > little bug fixes with your rewrite.
> 
> Thanks for the complement . I just needs a bit of time to hack around
> the quirks in all the zillion tray implementations.

Sorry, I didn't mean to be so harsh -- the new tray is really cool
(the hiding ability is awesome, and just being able to see a list of all
apps in the tray is useful too), and I imagine the code's a bit easier
to work with now, too. It's just a shame that Olivier and Jasper have
added little bugfixes and workarounds to deal with weird systray
clients that will have to be done all over again. I'm not a fan of
rewriting from scratch (in any situation) for just that reason.

> > Then maybe we need a new goal.  Really, the ideal thing is just to
> > stop using libxfcegui4.  We can just deprecate portions of it over
> > time -- it really doesn't matter in the long run, does it?
> 
> We will know in the feature, fact is that gui4 contains 80% of unused
> (by the plugins/core) code after the wnck merge, so it would be pretty
> stupid if we leave it there if you'd ask me. Deprecation is useful for
> libraries like gtk/glib etc, but not for a tiny lib that is used
> inside Xfce.

Possibly true, but it's really just eating disk space -- the parts of
the library that aren't used just don't get paged into RAM.  While it's
'aesthetically' yucky to have all that unused crap in there, there's
really very little cost.  Well, relocations, meh...

	-brian



More information about the Xfce4-dev mailing list