My weather app is different from others, because it tells you how long until the next snow. It only looks in the foreseeable of course, but if there is snow on the horizon, I get a snow icon, and that is the most important thing for me.
But the other thing I like about it is the probability graphs for precipitation. I like seeing the probability and the precipitation together – gives a more helpful guideline for making short term plans.
- It tells me how long till the next snow
- It gives probabilities for precipitation
After much research and lots of time spent on writing code for nothing, I think I have chosen which API to use. Well, two of them. See my findings and general (rather scruffy) notes here
One of the benefits of a blog is I can use it as a project manager. So here is the plan – kinda
Where we are now
An app which compiles – is slightly more organised than previously, and is pointing at the wrong API. Still not enough separation between data and controller, but Android does help to keep your view separate – so that is nice.
Where we want to be
Need to have a current weather page which shows what it does now
Need to have a dashboard – this might do better with gust information, and maybe get rid of info that is identical on the dashboard
Minutely graph needs to be changed because we have no probability. Perhaps we can have a rain and a snow graph instead
Perhaps we could have some indication of the tendency of precipitation on minutely, that comes in from the hourly – see what you can do.
Progress
Everything is now working up to the point where it was working before. There are differences in the forecast from forecast.io. But so far I am not unhappy with it. I believe the wind is more accurate in the VisualCrossing. So far the temperatures seem to be on track, but of course they are very easy to sway by putting a thermometer in sunlight. And forecast.io was never accurate on this front anyway.
I’m waiting for some decent rain, so that I can try to validate the graphs.
