Hi, just a quick heads up and a question:
On Monday 01 September 2008 03:31:25 Sisyphus wrote: > > To be frank I'm not sure anymore > > if I can fix this alone. It seems like a bottomless pit of wrong refcnts > > and > > trying to keep them all right is clearly expanding my mind... > > Mind expansion is generally considered to be a good thing ;-) It is a good thing. It's just that it can be a very painful one ;) But it worked. I was actually only one wrong refcnt away from a working module. A day to create a simple test case and two minutes to fix the bug using the test case... > I don't think Neil has time for *any* of his modules, these days, but I > could be wrong. Last I heard he was contactable at ActiveState - so if you > want to get hold of him to arrange maintainership (or any other matters) > try sending mail to neilw at activestate dot com. Didn't succeed in finding Neil, so I now got co-maintainership of Inline::Python and try to keep the module up-to-date. A little question for those with more experience in these matters: Python objects that get passed into perl space are wrapped in Inline::Python::Object objects. These have a DESTROY method that REFDECs the contained Python object, which is just as it should be. But additionally these objects have a magic mg_virtual->svt_free method which does the same. Clearly one of the two has to go, but which one? In other words: where does this kind of cleanup belong to? Thanks and kind regards, Stefan
signature.asc
Description: This is a digitally signed message part.
