BartlebyScrivener wrote: > Well, I tried to let this die because I just KNEW I was going to look > like an idiot before it was over. It's the .pyc versus the .py file. > Obviously I don't understand how that works yet. The .pyc file lags > behind the .py file? So when I run cbfindfiles.py I'm running the .py > version and when I run cbfindfiles it's running the .pyc version which > is not the same between edits.
I thought of that one, of course, but it can't cause exactly the trouble you describe above. If there's a .py in the same folder as the .pyc, it will not use the .pyc unless the timestamp encoded in it matches the one on the .py file (which, unless you go to extraordinary lengths, will never screw you up that way). On the other hand, if there was a .pyc in the current directory when you run it with "python cbfindfiles.py", but no cbfindfiles.py in the same folder, it will run the local one and not the .py which might be elsewhere, in your PATH. Any chance there are cbfindfiles.py* files in other folders or have you been doing this all from the same folder? > No need to explain. I'll go read up on how that .pyc file is generated. Don't think you look like an idiot (yet :-) ). Not only has everyone been caught out by this, sometimes repeatedly (ah... it happened to a "friend" of mine a number of times), but when you combine the weirdness of Windows PATHEXT and FTYPE and such, you can really get into awkward situations. I do hope you'll identify precisely what the issue was, so we can keep it in mind (if it's a new variant), or refresh our knowledge (if it's just the same old same old but we didn't realize it yet). Thanks. :-) -Peter -- http://mail.python.org/mailman/listinfo/python-list
