[Note: This problem is fixed in Python 2.5] I realized recently that I never posted an update about my piece on Python Memory Management. Basically, Python never frees any memory. I've written a patch that resolves part of the problem, and made a presentation at PyCon 2005 about it. It seems like there is interest in integrating this patch into a future version of Python. However, I don't have the time or the energy to constantly bother people until someone actually reviews it, so it has not yet been integrated. With my patch, Python can free memory, however there are many cases where it will not be able to do so, since if any object is still allocated in an arena, it cannot be freed. For more details, please see my PyCon 2005 paper [PDF] on the subject. On a personal note, it turns out that my PyCon presentation was one of the most popular! Wow, I'm flattered! I thought it was way too technical, but I guess I was wrong.