Spent basically the whole day trying to figure out why one of the two CMTSs is reporting huge numbers of offline modems even though you can manually ask any of the offline modems to scan and have it succeed.

It's one of those errors that seems to defy the logic of the application, the scanner doesn't even know which modem a CMTS belongs to when its doing the scanning, it only looks at the result after the scan has succeeded/failed (more precisely, it sends a message and the groups get the notification). How, in that situation, can one CMTS consistently be showing failures while the other works perfectly well?

There's lots of reasons, actually, but none that seem to apply in this case. It honestly just looks like a problem with the CMTS, but Bryan assures me the CMTS is fine, and the fact that individual scans work suggest he may be right. Very frustrating, all-in-all.

Worse, if my count is right, I only wound up logging 3 of the 6 hours I put in on it. Don't know where the rest went; very annoying to spend all that time getting frustrated and then realise you forgot to get the billing information recorded. Getting frustrated for free is not my idea of a good time.

The other hour of the day was spent on writing up and submitting 2 Firefox 1.0.1 bugs. They're pretty huge bugs, which makes me think they're likely due to AMD64 porting or similar effects, since I haven't seen a huge outcry about the release being fundamentally broken.

Anyway, my hands are on fire right now (tension and frustration are very bad for keeping shoulders and arms loose (to avoid muscle strain)), so I'm going to watch This is Wonderland before I do the evening's work on the PyCon paper.