Follow-up Comment #3, bug #25587 (project gnustep):
I know I argued that we should be defensive and leave the workaround
privately, but I'm swaying...
_undoManagerCheckpoint: will be called on each undo grouping except for the
first one (now after the fix to NSUndoManager). But we are processing changes
directly on begin and end groupings, and the bug was that the changes had not
actually propagated yet.
On the one hand I think it may be more efficient to simply register via
_enqueueEndOfEventNotification (except for the case that we are called by that
method).
Maybe we should spend some time with WO4.5 to find out when the checkpoint
notifications actually cause the editing context to process it's changes.
(And probably develop some test cases.)
So for now maybe revert and add a TODO... something like:
Figure out when checkpoint notifications should cause the editing context to
process changes.
_______________________________________________________
Reply to this item at:
<http://savannah.gnu.org/bugs/?25587>
_______________________________________________
Nachricht geschickt von/durch Savannah
http://savannah.gnu.org/
_______________________________________________
Bug-gnustep mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-gnustep
