> There were two other reports of failures in jump threading this morning, PR 8060 and mail from Argyrios on llmdev, which has a small testcase. Probably all the same thing. > I have also just confirmed that Argyrios' failure is unrelated. --Owen