Path: utzoo!utgpu!watmath!att!tut.cis.ohio-state.edu!ucbvax!decwrl!shelby!portia!Jessica!rick From: rick@Jessica.stanford.edu (Rick Wong) Newsgroups: comp.sys.mac.programmer Subject: Re: THINK C bug Message-ID: <4298@portia.Stanford.EDU> Date: 7 Aug 89 17:45:52 GMT References: <1989Aug6.155637.11270@jarvis.csri.toronto.edu> <2376@husc6.harvard.edu> Sender: USENET News System Reply-To: rick@Jessica.UUCP (Rick Wong) Organization: Stanford University Lines: 30 In article <2376@husc6.harvard.edu> siegel@endor.UUCP (Rich Siegel) writes: >In article <1989Aug6.155637.11270@jarvis.csri.toronto.edu> halpern@csri.toronto.edu (Charles Halpern) writes: >> >>This is the bug: when a project is running with the Debugger, and >>you hold the mouse button down, say while making a menu choice, after >>some random amount of time (usually 5 seconds to 2 minutes) the >>application will act as if mouse button is up (making an unintended >>menu choice, for example). >> > We've seen this one sporadically; it's extremely hard to reproduce, >and seems to be exacerbated by using INITs that themselves bottleneck >_StillDown. > When I'm using THINK C, I see it quite consistently. It happens in any mouse-tracking loop that uses a high-level Event Manager function (Still- Down, WaitNextEvent, GetNextEvent, and EventAvail). You may find it sporadic because your mouse-tracking loops don't last long enough. On my Mac II, it virtually always occurs within two minutes. If you use a low-level (OS) Event Manager function, such as GetOSEvent, the problem doesn't occur. It looks like the problem lies somewhere in the Debugger's patches to the Event Manager. I am using a "fresh" System 6.0.2, with NO INITs other than whatever might be in my System file. Rick Wong rick@jessica.stanford.edu