Richard Stallman <[EMAIL PROTECTED]> writes: > I verfied this with some trivial code that increments a counter. > AFAICT, the timer always repeats eight times. > > If timer-max-repeats is 10, but at most eight repeat calls occur, > that seems to suggest an off-by-two error in the code that checks > timer-max-repeats. Can someone check for this?
I have checked the code, and it looks correct. However, the purpose of the code (IIUC) is to only guarantee *at most* timer-max-repeats calls to the timer in the event of a long suspension (as a safety feature); it does not guarantee that there will be *exactly* that many calls. Off the top of my head, it's highly nontrivial to do the latter, because Emacs could be interrupted at any point, including the place where we are calculating when to call the timer next. So I don't think we should make any such guarantee. I suggest not making any changes to the code, and simply updating the docstring to properly explain the role of timer-max-repeats. _______________________________________________ emacs-pretest-bug mailing list [email protected] http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
