home.social

#waterwaymaporg — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #waterwaymaporg, aggregated by home.social.

  1. @hareldan for #WaterwayMapOrg I produce several X0 GiB GeoJSON files. (Well I make GeoJSONSeq which #tippecanoe can parse).

    I gotta say, GeoJSON is good for writers, because it's just text...

  2. @BUnicycling Hmmm. For #WaterwayMapOrg, i look at role `main_stream` so it doesn't matter whether they have a side channel or millrace role or not...

  3. @jtwcornell91 #WaterwayMapOrg (default map) does pay attention to canals in some ways, such as locks. That's how OSMers map them in a main river.
    WWM will also pay attention to OSM River Relations. So if "a river" is partially a canal, you can add that canal to the river relation, and WWM will treat it as a river.

  4. @amapanda

    Love it!

    It has been a while since I last visited your #WaterwayMapOrg site.

    I see changes that I'll need to make in order to include some large diversions in this area.

    I haven't edited OSM waterways yet, so this is now my New Year's resolution.

  5. RE: en.osm.town/@openstreetmap/115

    Thanks to the OpenStreetMap Foundation for paying for the hosting costs for #WaterwayMapOrg for the next year! 😊😊😊

  6. @pierzen I haven't changed anything on #WaterwayMapOrg. It works for me OK on Firefox v140 on debian trixie....

  7. @pierzen Canada is a mess on #WaterwayMapOrg. Yes, all the loops. But so many rivers not connected up.

  8. Reality (and therefore #OpenStreetMap) is fully of edge cases. Usually when parsing OSM data, you can ignore those, or flatten them down.

    The ““““fun”””” with #WaterwayMapOrg is that if that edge case occurs say... in a weird side channel of Mississippi, then you can't just ignore everything upstream (or downstream (depending)) of that point. Cause that's a large chunk of north america.

  9. My current #WaterwayMapOrg is re-working the “How are River A & B related? Tributary? Side Channel? etc“.

    Fitting OSM data into a small number of categories... what could possibly go wrong? 😈

  10. Updates on #WaterwayMapOrg should be working again.

    Sorry for the lack of updates. Some weird shell/makefile path error. I've hit it with a big hammer.

  11. I spent ages refactoring #WaterwayMapOrg code to produce debugging data to try to track down why WWM had the Missouri & Mississippi rivers separated by a weird fake river.

    Turns out there was a loop in #OpenStreetMap Data.

    Fix the input data, you fix the output data. 🙄🤦🏻‍♀️

  12. I have an idea for a new #WaterwayMapOrg feature. I wanna calculate the longest "source to mouth" length (and all the river names along the way).

    Then we can find the longest river system (as per OSM)! 💪🏻

  13. yay new #WaterwayMapOrg “group rivers by waterway _relation_” is showing the Rhein & Danube & Amazon and all those correct rivers.

    wait …

    why is it now reporting that the longest river in the world is Cooper Creek in Australia, at 15,000 km… 5× the length of the Amazon…

    /me goes back to the drawing board…

  14. awww yeah. 😎💪💫 After a few years, I've given in and added support for waterway relations in #WaterwayMapOrg code.

    Notice how the Rhein, Elbe & Danube are all connected up as a single feature with the same name? It's using the `name` tag from the relation to match them up.

    this is preview, WWM code will update haha soon enough. This'll mostly be useful for the River database list.
    #maps #geodata #OSMDataProcessing

  15. Turns out the “weirdly long lengths” on #WaterwayMapOrg was … because the hand written distance function I wrote had some sort of bug.

    I switched to a Proper Library™, and everything is Looking Sensible™

    en.osm.town/@amapanda/11447282

  16. Hmmm, Wikipedia tells me the #Yukon River is 3,190 km long¹, but on #WaterwayMapOrg, (taken from #OpenStreetMap), I have it at 4,924 km long²…

    I'm gonna have to figure out what's going on here. WWM's length is of the whole connected name group. Multiple connected side streams would increase the total length. But +1700 km?!

    (yes the length is haversine distance, no web mercator/geometry stuff).

    ¹ en.wikipedia.org/wiki/Yukon_Ri
    ² waterwaymap.org/river/yukon-ri
    #geo

  17. Just spent the weekend at Karlsruhe Hack Weekend at @geofabrik working on a cool new feature for #WaterwayMapOrg

    I had to set up CGI on Hetzner’s web hosting, and it felt like I time-traveled back to the early 2000s—pure old-school web vibes. 🕹️✨

    Now, fingers crossed that I can efficiently serve my millions of tiny files

  18. I am using an #LLM (#ChatGPT) to learn how various hydrological features are called. 😂😭

    “When river A flows into B, what do we call B from A's perspective?” etc 😂😂😂
    #WaterwayMapOrg

  19. One of our 2025 New Year's resolutions was to restart our very long running #OpenStreetMap blog interview series, and I am pleased to share our first interview of the year - we chatted with @amapanda maker of WaterwayMap

    blog.opencagedata.com/post/ope

    #WaterwayMapOrg

  20. Wanna help make #maps of #waterways from #OpenStreetMap? #WaterwayMapOrg now includes a first attempt at data export for you.
    read more: github.com/amandasaurus/waterw

    I'd like people to try it out and give me feedback on that data, how it's structured, etc. Is this the best format/layout?

    Share the maps you make please. 🙂

  21. If you see something that looks wrong on #WaterwayMapOrg, and looks correctly mapped on #OpenStreetMap, then you can open a new issue to fix how WWM calculates things. 😁 Is there a tag it should be using? etc.

    github.com/amandasaurus/waterw
    #OpenStreetMap #WaterwayMap

  22. I think I've hit the limits of what #tippecanoe can do. I use it to generate #VectorTiles for #WaterwayMapOrg. It's great for turning a geojson file into a #pmtiles file.

    The main maps on WWM.org are created from 20 GiB GeoJSON files. That's fine. 👍 But now I have a different data file which probably needs a lot of specialized zoom level processing to be usabe.