On Fri, Jan 12, 2024 at 05:17:02AM +0800, Zorro Lang wrote: > > No, it doesn't.. I actually did exactly that first, but that causes the > > test to be _notrun instead of reporting the error and thus telling the > > author that they usage of this helper is wrong. > > So below "usage" message won't be gotten either, if a _notrun be called > after this helper return 1 . True. But for the case reproducing my original error where I just misplaced it it does get shown. > If there's not _notrun after that, the message will be gotten I think. > So I think the "return 1" makes sense. As this point we might as well skip this patch, as it won't be useful.