I went to three sessions in the morning: Building Leafy Chat, A Short Pinax Tutorial and Import This, that and the other thing: custom importers. If you're interested, you can read on to see what I thought.
Alex Gaynor spoke on the "real time web" using AJAX and Comet (although the official title was "Building Leafy Chat, DjangoDose, and Hurricane, Lessons Learned on the Real-Time Web with Python". The first topic he spoke on was Leafy Chat, a chat program in the browser using django that he built in a team in 48 hours in a django dash competition. Leafy Chat uses JSON to pass around the data packets from the client to the server (and vice-versa). While it worked, it wasn't scaleable. The next topic was DjangoDose, which uses orbited, twisted and stomped. This web application worked better but wasn't completely stable either as it depended on Twitter and if Twitter went down, so did DjangoDose.
The next topic was Hurricane. It used jQuery, the Tornado server and multiprocessing queues to produce and consume data. It could do a twitter feed and chat, but the problem was that the application state was completely in memory.
The last topic was Redis. It builds on the previous examples and does much of the same things, just on a larger scale. See http://us.pycon.org/2010/conference/schedule/event/10/ for the slides. I didn't really follow how this was different from the other stuff he had created, but that's probably my fault,
The 2nd talk was by Daniel Greenfeld (AKA: pydanny). I think he was either the founder of the Pinax project or one of the major players. Unfortunately, he had serious technical problems that made following his talk rather difficult. He didn't even get to show off any code, so I was a little disappointed.
The last talk of the first session was by Brett Cannon, a Python core developer. It was rated as advanced and it was. I didn't really understand what he was talking about, so I left early to go to lunch. The second session was much better, so look for my post on that soon.
Copyright © 2024 Mouse Vs Python | Powered by Pythonlibrary