I think UmMiUme as a kind of platform where arbitrary providers can integrate their feeds (= channels). Since this would drive additional traffic to their own offering, it might be attractive enough to the information provider that they would pay something for it.
Based on this platform a client app (at the moment https://ummiume.mapresso.com) allows end users to get information about their location.
Channels and Items
A channel provides information on a specific topic. A query with a coordinate provides information about that particular point. In the case of continuous phenomena (e.g. temperature), you get one item with the value at that point. In the case of points distributed over space (e.g. Wikipedia entries), the result is a list of entries (items) sorted by distance from that point.
As sources I use open (and also less open) information on the web: GeoRSS, APIs, scraping of websites, files etc.. which I then harmonize to GeoRSS feeds. Of course, direct GeoRSS feeds would be ideal, but these are rare (hence the various earthquake examples).
Harmonising and Caching
Depending on the topic (“channel“) I fetch the information directly from the source, because they are different depending on the location or time. In other cases, I retrieve the data periodically and buffer them (as “items“) in a MySQL database (e.g. construction projects, airports, etc.),
Areas of nature (weather, geology, tides), culture (museums, monuments, events), infrastructure (public transport, shared mobility, authorities, mailboxes, …) and administration and politics (construction projects, construction sites, election results).
From very local, point-related info (temperature, coordinates, altitude, geology, etc.) to very wide ones (current distance to planets, asteroids and comets passing by Earth (“Close Objects”)). Currently, a lot of sources covers channels on a local or country level, using e.g. Open Data of the City of Zurich (https://data.stadt-zuerich.ch) or Switzerland (https://opendata.swiss/en)