On Tue, 6 Jan 2009, Daniel Fischer wrote:
Could you elaborate? I couldn't find an inconsistency using your previous code, it behaved as it should (until I ^C-ed it).
In several versions of the code, now unfortunately lost because of a crash on a power failure (which is extremely rare where I live), I did not get any "goOn" despite the value of gTst3 indicating I should, or where, according to your analysis, I should have gotten a single "fail," I didn't. If I can prod myself into recreating now lost code (unfortunate, sending out the wrong version of the code and losing the right one, only to replace it with one that seems to work now [yes, I'm careful about deleting .o and .hi files]), I'll do so, and report it as a bug according to the instructions in another post. Since I have workarounds and I am using a back-dated version of GHC, it probably isn't worth too much more attention, although I'll keep a wary eye open. Thanks to all for attention. Best, Murray Gross