It has to do with how nuke exposes itself to Python and how data is represented. It's not an uncommon way of doing things.
When you get a python object that represents something from nuke, like a node, it's only a soft link representation. Meaning if nuke deletes the node your python object is now like a broken symbolic link. The Python variable is only a pointer not the real thing. Knobs and nodes you should only treat the Python references as temp vars and not store them for long term referencing otherwise you will run into such issues. On Friday, August 28, 2015, Catherine Blanco <[email protected]> wrote: > I'm getting that a lot lately, too, and haven't found anything useful on > the interwebs either. Would love to make it go away. Right now I'm > resorting to ripping out chunks of Python code to try to narrow down where > it's happening. > > P.S. Hi, Den :-) > On Aug 28, 2015 2:38 PM, "Den Serras" <[email protected] > <javascript:_e(%7B%7D,'cvml','[email protected]');>> wrote: > >> Am I the only person plagued by the "A PythonObject is not attached to a >> node" error? There's very little about it on the Web considering how often >> it plagues me, almost like it's following me from studio to studio... It's >> not even consistent. A relaunch will make it go away. Refreshing a node >> will make it appear. Sometimes. >> >> Argh, >> Den >> >> _______________________________________________ >> Nuke-python mailing list >> [email protected] >> <javascript:_e(%7B%7D,'cvml','[email protected]');>, >> http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python >> >>
_______________________________________________ Nuke-python mailing list [email protected], http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
