Archives week 49 of 2004
Dec. 6, 2004 - Dec. 12, 2004
Got home from grams place (score: 2 all) and realised I had a bit of a fever, proceeded to develop all the classic flu symptoms (including the oh-so-lovely aches and pains in joints and a nice sweaty (and high) fever).
Luckily the fever broke around 10 or so, but then Rosey wanted me to review ...!-->!-->
Woke up late, did the week's grocery shopping, then took a nap (planning for just half an hour). Instead I woke up at 6pm, just two hours before the party. Decided to start working on my proposal for PyCon... got the summary and the like done and had just started working on the outline when ...
The demo instance has been crashing with memory errors after about 770 minutes of processor time, or about 12 hours of wall-clock time. As of now, we're up to 767 minutes, with only 153MB of memory used.
So often these leaks have tormented me; pretending to be gone, only to reappear ever-so-much slower than before. ...!-->!-->
Yet another leak discovered and fixed. Basically there was code that had started out storing a single object as the value in a dictionary. When checking whether to delete that object (a callback for a timeout), the code was checking to see whether the object it was about to delete really was the deferred object ...
So far the only failure that's been seen is in Numpy 23.6 (23.1 works), and it works with a different version of one of the packages. Was very pleased when I just typed "vc7" and was able to move both packages to Python 2.4 just by running setup.py again (having just finished building them for ...
I'm trying to release new PyTable and BasicPropery versions. Unfortunately, it's been something like 40 minutes, and I'm still waiting for the dratted SourceForge to load the release-editing pages. And the comics.com pages are taking... well... it's hours and hours anyway, I started it downloading when I woke up at 10AM, and it's still not ...
For those with a yen for PyGreSQL, there's a new version of PyTable that has preliminary support for it. It runs most of the PyTable tests, but fails miserably in the real-world project (some problem with the reverse-engineering of the schema from the database).
Anywho, there for those who want to play with it.
Wound up going down to Givex to work with D'Arcy on the EFT server. Spent most (almost all) of the time getting NetBSD packages set up for BasicProperty and PyTable, then getting a PyGreSQL driver written (and debugged) for PyTable (D'Arcy won't even install PyPgSQL).
Still problems with the PyGreSQL driver that I'm intending to ...!-->!-->
Well, that one leak definitely wasn't the end of it. Still seeing a leak of 4KB every 5 or 6 minutes when running the tests in a tight loop.
Blah. Should have gone out for coffee with Shane, would have been more productive.
File this in the "stupid things that waste half an hour of debugging time" category. Switching to trying to verify the memory leak fixed, I suddenly discovered that something like half of the tests were now failing... so combed through all of the changes trying to figure out what I screwed up... turns out I'd ...