On Tue, 12 Jun 2012 12:12:55 +0200, Alain Ketterlin <al...@dpt-info.u-strasbg.fr> wrote: >You misunderstand the problem here. It's not about the duration of the >actions, it's about the latency it takes to read/parse/execute the >script. HTTP is stateless anyway, so if the same "interpreter" handles >several requests, what you save by keeping the interpreter alive is the >load/parse phase. If you relaunch an interpreter for every HTTP request, >you pay the same price again and again for something which is not even >related to your scripts' execution.
Thanks for the input. But I read that PHP-based heavy-duty web servers compile the scripts once and keep them in a cache, so they don't have to be read/parsed/executed with each new query. In that case, what is the benefit of using a long-running process in Python? I enjoy writing scripts in Python much more than PHP, but with so many sites written in PHP, I need to know what major benefits there are in choosing Python (or Ruby, ie. not PHP). Apparently, very few people use Python à la PHP, ie. Python code embedded in web pages? -- http://mail.python.org/mailman/listinfo/python-list
