[Python-Dev] SVN rev. 85392 broke module handling in py3k (original) (raw)
Antoine Pitrou solipsis at pitrou.net
Fri Oct 22 18:17:55 CEST 2010
- Previous message: [Python-Dev] SVN rev. 85392 broke module handling in py3k
- Next message: [Python-Dev] SVN rev. 85392 broke module handling in py3k
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Fri, 22 Oct 2010 17:37:53 +0200 Stefan Behnel <stefan_ml at behnel.de> wrote:
I think the general problem is that "the module" can be a pretty broad thing, potentially referring to all sorts of stuff such as (usually) several other modules.
I wouldn't think of unloading modules as a general problem. We should support it for various reasons (including your use case), but if it takes a further garbage collection pass to get rid of all stale objects, then so be it.
It would also mean that each and every function will end up in a reference cycle by default that will require garbage collection for cleanup.
This is already the case, since most functions are in a reference cycle with their own globals dict (== module dict). It's just that the reference cycle would then include the module object as well, but I don't see it as a big drawback. The module object holds little to no state except its dict, precisely.
Regards
Antoine.
- Previous message: [Python-Dev] SVN rev. 85392 broke module handling in py3k
- Next message: [Python-Dev] SVN rev. 85392 broke module handling in py3k
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]