[Stackless] why use twisted with stackless?

Alec Flett alecf at flett.org
Thu Apr 12 01:44:30 CEST 2007


(more from me ... :))
So I was really surprised to find all the posts about Twisted + stackless
because I feel like if anything they are competing models for async
operations.

I'm in the process of trying to move a large, very synchronous codebase
built on a lot of WSGI middleware off of mod_python + prefork Apache (sorry
to keep repeating this, just trying to give context) and stackless seems
like an obvious solution because I can make minimal changes to the
underlying code - mostly I can just tweak the I/O boundaries and add the
occasional stackless.schedule() for cpu-hungry tasks. I'm basically just
using the Stackless HTTPServer that Richard directed me to, and diving
immediately into my WSGI apps.

But that said, I'm really curious what would drive a new project (like mine)
to Twisted. Stackless seems obvious and I've started down this path but I
don't want to overlook
anything if there's real value in Twisted that I'm not aware of. The only
thing I'm missing in my move from apache so far is logging, but that's what
'import logging' is for, right? :)

I attended some Twisted talks at PyCon last year, so I get the basic idea
(and it bears an uncanny resemblance to the network library in Firefox)

So mostly I'm just looking for opinions - what does Twisted buy you? Why
would I consider it?

Alec
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stackless.com/pipermail/stackless/attachments/20070411/c82c80c5/attachment.htm>
-------------- next part --------------
_______________________________________________
Stackless mailing list
Stackless at stackless.com
http://www.stackless.com/mailman/listinfo/stackless


More information about the Stackless mailing list