Day sucked into the abyss of pair-debugging (Where did it go? It was just here a few minutes ago...)

Day was rather fragmented (again). Mostly due to calls from D'Arcy. Was getting weird failures in the "EFT" server (now the "Payment" server). Turned out to be caused by a socket not getting closed in the test cases. Only showed up on the production server. Only occurred when the tests were run in given order. Produced essentially garbage error reports (DNS resolution failures, due (it turns out) to exhausting available sockets). Another one of those "the code is fine, it's the test harness that's screwed up" situations that always seems to compensate for any benefits in test-driven development might give you for non-trivial projects.

Also spent half an hour trying to figure out how to get Gentoo to build OpenOffice 2.0 candidate for AMD64 so that I could work on the Cinemon documentation on Linux. Eventually decided it just wasn't worth company time to work on it and booted into Win2K... at which point D'Arcy called and I had to boot back to Linux again... then emailing Bryan a few times... then trying to track down where the spec for the "billing system query" has disappeared to (it doesn't appear to be anywhere but in the code?!), and a dozen other little side-tracks each chipping away a few precious minutes from my day :( .

Anyway, I think I need to get some food, then maybe take a walk and clear my head. Tomorrow I'm going up to ACI to meet with Bryan (and maybe a few techs) to start hammering out 1.0 marketting and deliverables.


Comments are closed.


Pingbacks are closed.