gtg team mailing list archive

[Luca Invernizzi <invernizzi.l@xxxxxxxxx>]

This is caused by misplaced signal emitters.
The signal "task-deleted" was originally intended to be emitted after the deletion (unfortunately, in the code is "almost" after the deletion). The problems with tags is also due to misplaced signals.

Also, in the requester, the task_delete() function has an unfortunate name, since it's not made to delete tasks.
Tasks should be deleted with Task.delete().
That task_delete() function is called at the end of Task.delete() to proceed to the deletion of the task from the backend.

I'm mostly fixing these bugs in my GSoC branch, because I've changed
things here and there, so the fixes do not apply directly to Trunk (and
writing a fix just for trunk would create a bzr merge hell).


** Changed in: gtg
       Status: Fix Committed => In Progress

** Branch linked: lp:~gtg-user/gtg/multi-backends__invernizzi_gsoc

-- 
sometimes task is removed before task-deleted signal handler is executed
https://bugs.launchpad.net/bugs/582620
You received this bug notification because you are a member of Gtg
contributors, which is subscribed to Getting Things GNOME!.

Status in Getting Things GNOME!: In Progress

Bug description:
Signals are executed when "there is time for them", without specific limits.

Now, when we issue the signal task-deleted, we should do a series of things (updating tags in the tag pane, removing the task and its children from the main view...).
A number of times we need to access information about the task that has been removed. Example: we need to check if the tag that we are updating in the tags pane should be shown or not (are there still tasks tagged with that?)

This causes a series of problems, among which the drag and drop "phantom tasks" (tasks that are shown twice), or crashes upon deletion of tasks because the signal arrived too late and we cannot get the list of tags that need to be updated.

How to solve this?instead of passing the tid along with the signal, I think we should pass a copy of the whole task.
I understand there are a series of things that need to be well thought out also with this approach and, since it's late and tomorrow is GTG Regression day, i'll regress to bed.