Hi! El mié, 12-05-2004 a las 09:25, Richard Hult escribió: > On ons, 2004-05-12 at 07:41 +0000, Alvaro del Castillo wrote: > > Hi guys! > > > > I have added a patch to bugzilla in order to solve the bug: > > > > http://bugs.gnome.org/show_bug.cgi?id=141346 > > > > If everything is ok, I will commit it. > > You are leaking GErrors, I think. > I will check it. > We also need to be very careful when/if doing this change. > > Only the root of a subtree emitted the "removed" signal before, and the > rest of the code takes that in consideration, i.e. by traversing the > tree itself and removing canvas items, disconnecting signals, etc. If we > change it to emit the signal for the while subtree, we probably need to > change all of the other code. Did you test this patch with a complex > project with subtrees, links, resource allocations? > No, I haven't test it yet. I have the same feeling as you that this is a little but very dangerous change. Maybe, we can follow other approach: when a task is deleted, we can travel all the subtree with tasks and remove the relations in the hash in planner-gantt-chart. But I am not sure if the subtree will be alive when we receive the deleted signal. I will work on it and report you. Cheers -- Alvaro > /Richard
Attachment:
signature.asc
Description: Esta parte del mensaje =?ISO-8859-1?Q?est=E1?= firmada digitalmente