On Mon, 25 Sep 2017 23:19:46 -0700, brian.carpenter@gmail.com wrote: > Whoops, that should be "signed" not unsigned. The attached patch fixes this, and seems reasonable. The value is getting initialized to effective infinity, and then incremented. Just above, there is a special case to not incrment infinity, and it seems proper to do the same thing here. But since I don't understand study_chunk(), I won't apply the patch until the middle of April, to give people a chance to say that it shouldn't be applied. -- Karl Williamson --- via perlbug: queue: perl5 status: new https://rt.perl.org/Ticket/Display.html?id=132164