* Add dependency on the `thiserror` crate
* Add a global `Error` type, but also `maps::Error` and
`providers::combined::MergeError` for convenience
* Add matching `Result` types that default to the respective `Error`
type
* Refactor code to yield all kinds of error variants
* Add FIXMEs where library errors still need to be handled
* Remove documentation that explained why `None` was returned, this is
captured in the error now
* Move the function `draw_position` to the maps module and split it up
* Replace and refactor `pollen_at` and `uvi_at` methods on `Maps`
by `pollen_mark` and `uvi_mark`
* Drop the `pollen_project` and `uvi_project` methods on `Maps`,
just call the `project` helper method directly
* Add `map_at` and `mark` helper methods that handle maps slicing
and drawing
* Rename `pollen_sample` and `uvi_sample` methods on `Maps` to their
plural forms
* Also, rename the map handlers to `map_address` and `map_geo`
* Use the `Europe::Amsterdam` time zone from `chrono_tz` to determine
what date/time it is in that time zone
* Parse timestamps in the rain text API relative to this date/time
* Add the `fix_items_day_boundary` function to fix up stuff if
the series of timestamps in the rain text cross the day boundary
* Use `Position` everywhere instead of latitude/longitude float values
* Implement `Partial`, `Eq` and `Hash` for `Position` so it can
part of a cache key
* Drop the `cache_key` helper function
* Rename the `address_position` function to `resolve_address`
* Add methods on `Position` for formatting latitude/longitude with
a given precision (used for URL parameters in providers)
* Also cache address geocoding requests to OSM Nomatim!
* Use the `cached` crate for an easy implementation
* Add the `cache_key` helper function to deal with floats being annoying
* Cache Buienradar get request for 5 minutes (per position/metric)
* Cache Luchtmeetnet get request for 5 minutes (per position/metric)
* Note the `Item` structs need to implement `Clone` now because
the cache will own them and Rocket will want a copy too
* Add the `csv` crate as a dependency
* Use the `Row` struct as intermediate object
* Turn the `parse_value` function into the `convert_value` function that
cannot fail