Palmer's Ephemerides

One of my backburner projects is an ephemerides service which will tell one such things as the sunrise and sunset times, phases of twilight and the moon. The script I'm using outputs its times in UTC, so if I wanted to use local time, I'd have to map from latitude and longitude to time zones, which would be an interesting task.

But I was thinking about the Sommerzeitwahnsinn and wondering why I should bother converting for my own timezone, which is so close to UTC anyway. Why shouldn't I just set all my clocks to UTC, and ignore BST this year? I'd run the risk of miscalculating the time that the rest of the Britons are on and being late for appointments and so on, but on the other hand I'd be able to calculate log time differences more easily, and I wouldn't have to munge any other times that were given in UTC, as for example the ephemerides script.

I'd also like to include weather forecasts, and shipping forecasts for those areas on the coast, since most weather forecasts don't come in easily repurposable and beautiful formats. But this would require mapping from coordinate data to location, and would probably be quite tricky since BBC Weather, at least, requires place names or postcodes as its input.

This notion of trying to encourage and proselytise repurposability stretches back locally to a discussion that we had on Swhack on 2006-01-22 that I haven't written up yet about doing something with the Semantic Web; and globally to TimBL's XML 2000 Talk wherein he says that "if you say what you mean rather than what you want done with it, you can repurpose it so much better"; something that became somewhat of a mantra for William Loughborough.

Sadly, United Kingdom government data is still very expensive. The Met Office charges; Ordnance Survey too unless TimBL is heeded. The problem is that these departments have to be self-funding, and have difficulty breaking even. As Tim commented: "The OS have always sold maps for money. That doesn't mean they have cut a net profit -- the impression I had (possibly wrong) was the government would like them to break even." The Oxford English Dictionary has a similar problem, though at least the first edition is slowly becoming available online, thanks to Kragen Sitaker.

If there's much left for the amateur ephemeridist to do, it's probably in this area, though there may be some small amount of experimental work to do that I'm not keyed into. For example, it's known that the phase of the moon is correlated to the intensity of aurorae, but that's something that was difficult to prove before it was objectively measured due to the fact that less aurorae are visible anyway when it's a full moon.