Tobias

Back in 2020 @waysidemapping@mastodon.social explored new UI ideas for the #OpenStreetMap iD Editor. Those explorations where very interesting but not continued. This weekend I re-deployed his old branch and rebased it on top of the latest iD version with LLM-support. You can explore and use it at https://deploy-preview-8--tordans-id-experiments.netlify.app/#background=Berlin-2019&disable_features=boundaries&map=19.94/52.47609/13.42093. Code at https://github.com/tordans/iD/pull/8 (rebase) and https://github.com/tordans/iD/pull/9 (original).

The @panoramax@mapstodon.space project created a AI model that recognizes German Traffic Signs on Street Level Imagery https://huggingface.co/Panoramax/classify_de_road_signs. Now we "just" need a German #Panoramax Instance https://www.osm-verkehrswende.org/root/posts/panoramax-bedarf/ to get this data into a Pipeline that helps keep OSM updated (like this https://radinfra.de/posts/2025-verkehrszeichen/).

@heigit@mastodon.social did a great comparison of OSM land use data with an external dataset. But I have a gripe with how they present OSM as the inferior dataset. I blogged about it at https://www.openstreetmap.org/user/tordans/diary/407524

RE: @yutannihilation@fosstodon.org

Looks like I should learn more about #DuckDB and how it might be useful to work with maps data from #OpenStreetMap. Any related OSM projects/explorations I should look into to learn more?

Wir suchen eine Next.JS Entwicker:in bei FixMyCity, wo ich arbeite.
Möchtest du dabei helfen, Software zu entwickeln, die die Planung und Umsetzung von Rad- und Fußverkehrsprojekten beschleunigt? => https://fixmycity.de/jobs/

@cantences ja https://www.openstreetmap.org/note/5008854