I found what seems to be an awesome Python parallelization library which gets around the GIL by using processes and IPC. The same mechanism lets it support distributing jobs to clusters across a local network or the Internet, with dynamic load balancing.<p>Sounds pretty great, right? However, I haven't heard of it before, nor have I been able to find much (recent) buzz or reviews about it. Any of you all have any experience with this library?<p>(link: http://www.parallelpython.com/)
I stopped doing much with python about the time I first heard of it. It has been mentioned here before:<p><a href="http://news.ycombinator.com/item?id=654842" rel="nofollow">http://news.ycombinator.com/item?id=654842</a><p><a href="http://news.ycombinator.com/item?id=223972" rel="nofollow">http://news.ycombinator.com/item?id=223972</a><p><a href="http://apps.ycombinator.com/item?id=147614" rel="nofollow">http://apps.ycombinator.com/item?id=147614</a><p><a href="http://news.ycombinator.com/item?id=107221" rel="nofollow">http://news.ycombinator.com/item?id=107221</a><p><a href="http://news.ycombinator.com/item?id=81707" rel="nofollow">http://news.ycombinator.com/item?id=81707</a><p>Perhaps there's less need for it since a similar module (multiprocessing) is now part of the python standard library.