Richard Stallman schrieb: > I agree that one-shot timers are generally better. But when I tried to > implement a one-shot timer for delayed autoselection, it occasionally > didn't trigger for some reason. Maybe I should give it another try. > > Please do give it another try, because if reactivations get lost > (other than due to an error occurring earlier in the timer function), > that would seem to be a bug in Emacs. If there is such a bug, we > should track it down. But first let's see if such a bug currently > exists. > > Can you do that, then get back to me?
In the meantime, I was able to make this work with a one-shot timer that reschedules itself. Hence, I think there's no such bug. FWIW, a possible sequence of events that triggered the problem on my system goes as follows: The timer function runs `window-at' which, on Windows, seems to break when the mouse is on a scroll-bar or on a menu item. Wrapping that `window-at' in a `condition-case' resolves the problem for me. Hence, apparently, the timer was not rescheduled due to "an error occurring earlier in the timer function". _______________________________________________ emacs-pretest-bug mailing list emacs-pretest-bug@gnu.org http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
