-
-
Notifications
You must be signed in to change notification settings - Fork 347
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Import new area (around Strasbourg - France) #1006
Comments
Sure! Send a geojson.io polygon please. You can also try the import yourself (https://a-b-street.github.io/docs/user/new_city.html#the-easy-method), but use the downloaded version (https://github.com/a-b-street/abstreet/releases/tag/v0.3.34) for that |
Thank you, I've managed to import a neighborhood into the downloaded app from a geojson area and started to discover the LTN part of the app. For the time being, I think I'm good importing in the UI every time I want to use my area. I may ask later for a map of my area to be included in the UI, depending on how other bike activists plan on using it locally. |
Awesome, keep me posted how it goes! Depending how dual carriageways and bridges/tunnels in the area work, the default boundaries might be really weird and adjusting could be tough, let me know if you hit problems |
Found this old issue : not an issue anymore, importing is fine for now. |
Re-opening this because we are starting locally to use the tool (LTN module) more intensively with bicycle activists and local authorities : more (non-tech) people will need to be able to manipulate the area, so the manual import will not be enough, also we will need to demonstrate and use the web version with local examples. Here are 3 json files for the area of interest. I'm wondering if the files cover an area of reasonable size to integrate in the tool ? I've downloaded the multipolygon version from geojson.io, but also split into 3 files as I didn't remember what would be best to implement. |
Sorry for the huge delay here! Here they are:
I had to adjust all 3 to not cross the German border. The import process finds the smallest extract from Geofabrik containing each boundary, and so if you cross a country border, it pulls down the Europe-wide file, which is too big. Is this OK, or do we need to fine-tune the boundaries? Some of the files are large to load in the web tool. They'll work, but be a bit clunky. Some of the auto-detected boundaries are also a bit broken. Please try things out and let me know if there's any issues we can improve. Exciting to see this get used! |
FYI: if it helps I sent PR #1126 to add the relevant GTFS. |
I beliebe it's ok to implement this way. Just to mention it, i am indeed experiencing some hiccups with the auto-detected boundaries Visually it seems similar to #1087 |
A new release has these maps: https://play.abstreet.org/0.3.49/ltn.html?system/fr/strasbourg/maps/center.bin There hasn't been recent work on the auto-detected boundaries, so I'm not surprised there are more problems here. The only news is that I started prototyping a very different approach to the LTN tool. Setting boundaries is hopefully simpler. Still cases with bridges and tunnels are a bit weird, but I have more confidence we can solve them with the new approach. You can preview the new one at https://dabreegster.github.io/ltn, but it's very early days still |
Unfortunately you're hitting the biggest unsolved bug with the LTN tool. It's probably not an OSM bug, just a problem with osm2streets-generated geometry, bridges / tunnels, or the "blockfinding" algorithm. I'm not sure there's a good workaround right now. The good news: I'm in the middle of rewriting the LTN tool from scratch, fixing lots of the big limits. (https://dabreegster.github.io/prose/dec2023/pt8_tech.html talks a bit about those problems.) The new version is very early, but we could at least see if the boundary you want to draw works with the new approach or not. Could you please try this:
One big difference between the two tools is about trying to auto-detect neighbourhood boundaries. Because auto-detection is so problematic in many cases, the new version doesn't do it at all. Instead, the user just draws the polygon they want, and hopefully the auto-snapping tool makes this fast enough. You can trace along "main" roads on all sides, or not, depending on the specifics of the area. |
Again, it's exciting so see what's coming (including the new walking-focused tool) It looks fine and I didn't detect any issue in the process : I'm impatient to see this new version grow and evolve! |
Awesome to hear it works, and it's showing up for me too: Very useful to know the new version can handle this boundary that the old can't. I'll prioritize development of it as much as possible. And you reminded me that sharing the proposals from the tool easily is something to consider, so filed a-b-street/ltn#13 |
Is there an easy way to follow the progress of the new revamped LTN tool ? (and to contribute with feedback, comment, translations, etc) |
I've only been able to spend a day or two per week on it, so it's felt quite tough to get it to some kind of MVP. How about as a simple improvement, I'll start a new GH issue in that repo to summarize recent changes? You can feel free to comment there or raise other issues. There are big UX / design questions I have that I can ask, and if you have opinions or want to get more involved and sketch a wireframe or something, that could also speed things up! Translations is a very good question -- it'd be better to get that started sooner or later, and I'm hoping it's much easier, since the frontend is now just a web app using a popular framework. Let me start those GH issues and I'll CC you in... |
Thanks. I believe we can close this issue now. What do you think? |
Would it be possible to integrate the area around Strasbourg - France ?
Planning to work on LTN locally and I'd love see what abstreet can do and provide feedback
The text was updated successfully, but these errors were encountered: