Description
Demistifying GIL - what's that, how it really affects performance, what changed in the last few years and what can we do about GIL in everyday programming.
There is still a lot of misunderstanding regarding GIL, coming from not-so-perfect documentation - not explaining the implications thoroughly and in a simple way; it's also often outdated.
Also, GIL's implementation changed a little in CPython 3.x that may affect programmers' choices.
This misinformation may, in some cases, drive them away from CPython where it's not really the worst choice. By providing comprehensive guide to today's GIL in CPython, this presentation aims to help make better design choices - be it choosing a concurrency/parallelism model, Python implementation to use or even whether or not to drop Python entirely.
There have been (and still are), of course, attempts to get rid of GIL and find another solution for the problem it solves - newest of which is Larry Hasting's project called GILectomy, which I hope to shed some light on.