[rrd-users] How store wind vector data?

[Simon Hobson]

Alex van den Bogaerdt <alex at vandenbogaerdt.nl> wrote:

> If you would use RRDtool, and do not carefully avoid all its consolidation, 
> you will end up with an average wind direction of 180 built from 350 and 010 
> at some point in the future.

Yes that's a tricky thing to work round.
<cough> years ago, with a different hat on, I had involvement with stuff I can't talk all that much about - lets just say that submersibles have a lot of similarities with aircraft when it comes to controlling them. There was talk about an autopilot on a previous project that wasn't designed to cope with the 359-0 problem, and while out on trials, while following a track close to north, it suddenly decided it needed to do a loop to get back on course. Ie the error had taken it across this discontinuity so the autopilot thought the error was 350+˚ instead of just a few ˚ :D
On our project, we made darned sure we tested for that during acceptance trials.