Django sprint wrap (for me)

Published: September 16, 2007. Filed under: Django.

I don’t know whether anyone else is going to continue on for the rest of the night, but I’m exhausted and I’ve got to be at work in the morning.

Today, for me, consisted of reading every open ticket in Trac and doing as much triage as I could; a lot of duplicates got closed, a lot of similar issues got consolidated, a lot of ancient tickets which have either been fixed or invalidated were marked as such.

Some highlights of the last 36 hours:

And the big one: sessions now use configurable backends for persistence, with a file-based store and a cache-based store included in addition to the original database-backed one. There’s still tuning to be done there, but once it’s humming along this will allow high-traffic sites to see some significant performance improvements by moving away from the DB-backed sessions store.

Fredrik had a script scraping Django’s Trac regularly throughout the sprint, and posted these final numbers when I told him I’d triaged the last open ticket:

We really ought to do this more often.

Post fun sprint stories in the comments if you’ve got ‘em, or your favorite improvement/bugfix; with that many changes going on, I know there’s interesting stuff I didn’t see as it flew by on the timeline :)