Author archives: Mike
Sat down and did the review and final edit of the metaclass presentation this morning before work. I have more hope that it will fit in the time-slot. Basically if I only hit the top-level bullets and gloss over the sample code we'll be fine.
The presentation final draft is due Wednesday. There's no business cards for Mikey yet. Work on importing framework must continue, as must the whole-system simulator. Too much to do, too few days. Oh well, off to it.
Apropos to the question of play in design, particularly the effect described in my thesis of providing a safe environment for exploring ideas:
What you see here is a fairly deep topic; getting caught up in trivialities and as a result losing the chance to demonstrate affection. It's something that's painful for many people, yet ...!-->!-->!-->!-->
Well, as everyone and their brother has been reporting, the DARPA race across the desert wound up with best showings of 7 miles (14km or so). Lots of coverage of the mechanical exoskeleton for soldiers as well.
Of course, coverage of Spirit and Opportunity has been phenomenal Still, they're expected to drive maybe a few ...!-->!-->
Went to bed right after work yesterday. Slept most of today. Yet still my eyes are sore and bloodshot. Keep getting up every few hours and trying to do a little work only to realise I'm not really in any shape for it. Suppose there's tomorrow.
Darryl has third in series of articles in his blog on using the Round Up API directly that I should take the time to read some day soon. Probably should have Tim read it as well, since the issue tracker is mostly his baby.
Well, it's finished. At least, everything is now reviewed and revised, with the code samples in colour-coded glory. Unfortunately, there's now 74 pages.
Got an email from customer' tech asking me to review the logs and see if there was a problem on our end during a given period. So, I check, which takes an hour or so of poking around in various log tables on various running instances, and then start drafting email to say basically "no, ...
So either I'm getting too deep and forgetting the audience, or the whole of the presentation is starting to make sense. Still need the section discussing how to decide when to use them, and a review of how metaclasses fit into the goals of meta-programming. It's fairly clear IMO, however, so yay. It's a much ...
Getting there on the full-plant simulation front. Discovered a serious bug in the OIDStore implementations in TwistedSNMP this evening, they were storing with naive alphabetical order when the requirement is for numeric ordering. There's a new release on SourceForge.