gtg-contributors team mailing list archive

[Lionel Dricot <ploum@xxxxxxxxx>]

Le dimanche 19 septembre 2010 à 01:05 -0700, Luca Invernizzi a écrit :
> Maybe we've missed some important bug. Explanation: I was trying to
> find the problem with gtk by disconnecting completely liblarch_gtk
> from the underlying tree (by disconnecting the treemodel).
> I figured that doing that I should come up with an empty tree but no
> complaints from gtk (since we were assuming that the errors were
> coming from threads and race conditions).
> Turns out that liblarch crashes also with the treemodel disconnected.
> That is great, because the bug might be easier to find. Any idea why
> is that happening? Is liblarch_gtk communicating with liblarch only
> via "node-added", deleted, modified and  reordered?


That's great, indeed.  Liblarch_gtk is of course also communicating with
liblarch with all the other functions in viewtree !
zjnmmmmmmmmmmmmmmmmmmmmmmmmmmmmm 

(oops, the cat walked on the keyboard.)

Last but not least, some liblarch.TreeNode function are still overridden
in task.py (which should be removed completely, one day or another).


If you can reproduce any bug without threading, that would be insanely
great and I commit to fix them quickly.