imports don't usually take up too much memory. I'd guess that even importing a very large number of modules isn't going to grow your process size by more than 10 MB.
imports are only evaluated once. So if you nest the import into a view, your process will be smaller until the first time that view is called. Then your process will grow in size depending upon the size of the module(s) imported. Repeat calls to that view won't grow your process size though. So you could get a small, temprorary decrease in footprint by nesting imports into views. But overall, it's likely to be a micro- optimization. Paying attention to the number of model objects created and used in a view is generally going to be much better area to focus on if you want to reduce your memory footprint. e.g. loading a lot of large models could add 200 MB to the process size. delaying imports for a little while could only delay growing the process size by 2 MB. But, if you want to get really fancee, you could use the zope.deferredimport distribution to define all your imports at the top of your source code, but defer the triggering of the import process until an object from that import'ed module is actually used. But this distribution wasn't written specifically with optimizing memory footprint in mind, but rather to allow faster start-up times and to trigger deprecation warnings when variable(s) in a module are used. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to django-users@googlegroups.com To unsubscribe from this group, send email to django-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/django-users?hl=en -~----------~----~----~----~------~----~------~--~---
