Closed hmatuschek closed 2 days ago
Yeah, the Repeaterbook API is broken right now too. ARRL says it gets its data from RFinder, but they're in flux right now--most of their assets, apparently including the directory, have been sold to a new company Unified Radios. The directory is online here: https://directory.unifiedradios.com/ but Rfinder's directory wasn't free and I doubt this one will be for long.
Despite RepeaterBook's shortcomings, Garrett seems like a nice guy, and I don't think the terms are too bad. Are you planning on removing it as a source, or just adding others?
I would try to keep the repeaterbook interface and add the other ones as primary sources. However, these sources work very differently. While I need to query repeaterbook all the time and thus increasing their load, I only need to fetch a few MB from the other sources once a week or so. Being so restrictive actually increases the load on the repeater book side. However, if it gets too difficult to keep repeater book in, I will throw that source out.
One additional source to consider is RadioID, although it's only DMR repeaters. See: https://radioid.net/database/dumps. rptrs.json is their entire repeater database and map.json is the subset that includes lat/lon coordinates. One downside is that they include quite a few repeaters that aren't on the air, but data quality is an issue with all the sources I've seen.
Okay, thanks for the info. I've added these sources (hearham is disabled for now, but can be enabled in the settings dialog). I kept repeaterbook but there is a server issue for non-US repeaters. You can test it in the branch https://github.com/hmatuschek/qdmr/tree/494-rework-repeater-database-sources.
It works for me! (I didn't test Repeaterbook's rest-of-world database because I know the endpoint is down.) One minor glitch: If I select or unselect sources in Settings, the changes don't seem to take effect until I restart, at least for the non-Repeaterbook sources. Perhaps this is because you cache them at start-up?
I merge the cached repeaters at startup into a single dataset. It is hard to remove some information from that merged set afterwards. So, I would keep it this way for now.
Repeater book is a really bad source. They have a very restrictive usage policy while relying on the community to keep their product up to date. I believe, that a reliable, community-driven repeater database with a liberal usage policy is a basic infrastructure that we all benefit from. Maybe someone at the large HAM clubs (ARRL, RSGB, DARC, ...) feels responsible to provide such a platform. Anyway, I need a replacement for the repeater book source. I'll switch back to