Ways to improve weather accuracy

Hello everyone,
I have noticed that the weather app isn’t capable of showing relevant weather information. In my case the nearest implemented station is 112 km (70 mi) away.

Germany’s meteorological service (DWD) has this list of CLIMAT stations, which gweather currently uses - https://opendata.dwd.de/climate_environment/CDC/help/stations_list_CLIMAT_data.txt. With those stations, the nearest location for me is 18 km (11 mi) away. This list alone can close about 40 open issues in libgweather. (Note: the stations don’t fulfill the 100 thousand inhabitants rule currently active)

The DWD, also provides more locations in Germany, using brightsky.dev, which simplifies access, I am able to get forecasts of 5,39 km (3,35 mi) distance, which is more than enough.

There is also the Locationforecast of the Norwegian Meteorological Institute, which returns a forecast on a grid of 0.1° x 0.1° lat/long (roughly 11.111.1 km/ 5.45.4 mi).

Also, why are weather locations saved to an local file in the first place? Accessing weather data already requires an internet connection. All through it would slightly impact the user experience, requiring the users to enter an location with longitude an longitude seems to be easier to use with providers

I have a bit of experience with python and would try to contribute as a newcomer, implementing this looks feasible for me. But since I have no idea where to discuss possible changes or where to get approval I’ll post it here.

If you want to implement a new weather provider for libgweather, feel free to do so; you can look at the existing providers:

The provider will be used by gweather_info_update().

Because the location data is not used just for weather data: it’s also used to compute things like sunrise and sunset time for the night light; it’s used to get the time zone of a location for the world clock; and it can be used for anything that requires location information. These services do not need a network connection, and using a local file allows them to access the data more quickly and more efficiently.

This is really not an acceptable UX, sorry.