Today, there are two types of roads colored pink.

This is a gated/private community in Seattle called Broadmoor. There's a physical gate surrounding this neighborhood; you have to show an ID to a guard to walk or drive in.

This is a Stay Healthy Street in Seattle. Also called "play street", "living street", "low traffic neighborhood" in different parts of the world. People can drive through there only if they live on that street, or they're making deliveries. Pedestrians and bicycles and people playing outside are supposed to be the main use. They usually have some kind of sign or barrier at the entrances:

We should probably depict these differently. The pink color isn't meaningful at all. We have the unzoomed view, shown above, and also the zoomed-in detail view like this:

Maybe we could have a simple 2D top-down rendering of a barrier or a gatehouse? For the unzoomed view, I'm not sure what to do.

@trangrei

For issue #813

Problems:

• ~~Currently crashing in map_model::objects::traffic_signals::Stage::max_crosswalk_time failing to find a movement in i.movements~~
• I added PathStep::ContraflowTurn mostly by adding it to the enum and then fixing all the non-exhaustive pattern errors, which means I likely missed some matches which use _.
• I haven't touched pathfinder v2 yet, notably into_v1_walking would likely generate TurnIDs which do not correspond to any Turn

other TODOs:

• Degenerate intersections still have two crosswalks (they had four before?). Do they not have SharedSidewalkCorners?
• Once all duplicate crosswalks are not being created, remove other_crosswalk_ids and its one reference.

see https://github.com/dabreegster/abstreet/issues/154#issuecomment-658026415 for current plans

see https://github.com/dabreegster/abstreet/issues/205#issuecomment-660707593 for warp hint

sim/src/make/generator.rs contains code for generating trips

See https://github.com/dabreegster/abstreet/pull/183/files for related code

See https://github.com/dabreegster/abstreet/commit/ead4a99682d11ec4af4824a1ab4c3854bae7b0d4 for related code

I have no idea how "random" traffic generation works, but currently it results in poor results for large maps.

Initial traffic at midnight seemed reasonable (maybe bicycle traffic was too high and bus traffic was too low), but later instead of ramping up it started dropping!

I expected game melting down/crashing on attempting to simulate tens (hundreds?) thousands of active trips at once rather than drop to zero active trips.

EDIT: yes, it is a prime example of not reading warnings. Random is described to behave in this way - see https://github.com/dabreegster/abstreet/commit/da1731ccb2d549ed95516de76881ab53f67e8dd3#diff-628ee6879bfcdd889804ae5371c9a26cR223

This is a followup to @westnordost's suggestion in #704 to draw a more explicit curb between the road and sidewalk. I incorporated a few more ideas here too, pulling lots of colors from Streetcomplete (https://user-images.githubusercontent.com/7988080/103153247-827c1f00-478f-11eb-9d1f-4d12017533aa.png) and also its map style (https://raw.githubusercontent.com/ENT8R/streetcomplete-mapstyle/gh-pages/images/light.png). I'm really liking how things are looking and am considering a big color overhaul based around this. Any initial feedback? @Robinlovelace, @michaelkirk, @trangrei

Some of the changes:

• Using Streetcomplete colors for the map background, water, parks, unzoomed roads
• Drawing a bold line between the road and curb. (There are some bugs near intersections that I'll fix if we decide to push forward with this style.)
• Use Streetcomplete color for zoomed-in lanes
• Stop coloring intersections and parking lanes a different color. It usually doesn't match reality; just show paved surface, and add markings on top of that.

If we go forward with this, there's lots of code simplification I'll do, and rethinking the styling for bike/bus lanes, access-restricted roads, night mode, etc. Also need to make sure layers and agents look good on this scheme.

We could also just incorporate a few of the ideas -- the color changes and adding the curb could be independent steps.

Before: After:

Before: After:

Before: After:

…cted bike lanes. They're only created manually right now, to explore rendering.

The goal is to start modeling different types of bike lanes -- some are right in the door-zone of street parking, others have flimsy little flex posts protecting from cars, but some have proper barriers in place. Eventually we'll use this in the routing and simulation layer. But the goal for this PR is actually just to explore the rendering style.

I've started with 4 types of barriers, but happy to add more!

For each of these, any better ideas for visualizing in our top-down view?

Stripes

Top-down rendering: Reference:

Flex posts ("wands" in the UK)

Rendering: Reality:

Planters

Rendering: Reality:

Jersey barrier

Rendering: Reality:

Next steps

There's more work needed to create/modify these properly in the road editor UI. I also want to start tagging these in OSM, and then importing into A/B Street. https://wiki.openstreetmap.org/wiki/Berlin/Verkehrswende/Radwege#Tagging-Beispiele seems to have some traction.

In collaboration with the University of Manchester, we're trying to initially apply the LTN tool to study the Levenshulme LTN. Urban Observatory has sensors measuring air quality, pedestrian counts, and vehicle counts in the area. There should be data both before and after the LTN was put in. So, the goal is to take the LTN's impact prediction about short-term detours, and compare it to the real observations.

What exactly is the LTN scheme?

We'll need a map representation before and after the scheme was put in.

• [ ] Find a definitive boundary
• [ ] Find the location of the filters
• [ ] If recent OSM edits have picked up the LTN, transform the small bike-only segments into filters, so they can be manipulated in the tool

From https://s3-eu-west-2.amazonaws.com/commonplace-customer-assets/levenshulmeandburnageactiveneighbourhood/FAQ%20Design%20Detail.pdf:

https://secure.manchester.gov.uk/info/200024/consultations_and_surveys/8009/levenshulme_and_burnage_active_neighbourhood/3 lists 14 filter locations

When were the filters installed? (And double check -- do we have counts from before that?)

Hi, your progress with the LTN Tool is exciting and it's already potentially useful. London UK is a focus of a lot of effort in planning LTNs.

As you may know, London is administered as 33 boroughs that act as separate cities for most road planning. My own borough is Camden https://www.openstreetmap.org/#map=13/51.5428/-0.1809. It would be great to get the OSM data for Camden imported to AB Street. I'd be happy to designate the area or to download the relevant section of the OSM database.

Once knowledge of the LTN Tool spreads I imagine there will be demands from quite a few other London boroughs too. Does it make sense to import the OSM data for all of Greater London (http://download.geofabrik.de/europe/great-britain/england/greater-london.html)?

Separate cyclepaths like https://www.openstreetmap.org/way/54563641 excluded; let's try enabling them.

https://www.openstreetmap.org/way/4188078 has cycleway:left, but it winds up on the right -- I think the left-hand driving side needs some fixes here.

I built the simulation game from source. After starting the simulation, i see a mix of readable text and non-readable pixels, where i assume, another text should be visible. Is there a missing font or do i have to define a prefered language for the game?

What's involved with this? It's not just converting the OSM data right? Are locations hardcoded in somewhere? My city has terrible street traffic. I'd love to modify this to support other cities easily.

Hi,

great project!

I tried to import my hometown, Münster in Germany. I used the (unzipped) .osm file downloaded from bbbike.org. I tried to import this with the following command:

./import.sh --oneshot=Muenster.osm

The import fails with this error.

thread 'main' panicked at 'No bus route from Position(Lane #102018, 27.2693m) to Position(Lane #86195, 7.4311m) now for Bus 329: Sendenhorst, Realschule => Hoetmar, Hagenhoff! Prevent this edit', map_model/src/pathfind/walking.rs:329:17

This might be related to #190, but I didn't investigate any further beyond this error message. I got the same error with a manual bounding box download from bbbike.org, though.

Currently the raw_map layer embeds osm2streets StreetNetwork, just adding A/B Street-specific stuff in addition (buildings, areas, transit routes, and things like crosswalks & modal filters that aren't yet in osm2streets). But then the map_model layer, which is used by the traffic sim and all the apps, has its own representation for roads and intersections, and all the osm2streets stuff is gone. This was OK to start, but it's starting to get a bit tricky:

• A/B Street edits lane specs and needs to regenerate geometry and movements, so we have an adapter kind of hack around it now
• Rendering code is starting to be duplicated between map_gui and osm2streets. And then again in #1038 for Bevy
• There are some osm2streets API concepts like RoadEdge that'd be useful to use over here now, but they're not easy to get at

I'd like to explore options for embedding StreetNetwork inside map_model::Map too. Reasons to be wary of this:

• Keeping serialized file size small and loading times quick for large maps is important, potentially at the expense of code readability / API niceness
• the API surface of Lane, Road, Intersection is huge and messy
• Keeping some indirection / wrapper between osm2streets and here is a good idea for evolving the two codebases at different speeds

Another vague idea is to embed osm2streets::{Road, Intersection} structs inside of A/B Street equivalents, and keep a few extra fields over on this side

CC @BudgieInWA

Show the OpenStreetMap community what data you are using. Add your project to the list of Taginfo Projects.

(you may provide multiple, i.e. if there is different themed layers. or -if possible- for Streetmix)

Platform: Linux (ArchLinux, Gnome 43.2, Wayland)

When starting the game with RUST_LOG=trace RUST_BACKTRACE=1 cargo run --bin game -- --dev, I observe the following behaviour:

• the window starts with a 800x600px size, then goes full screen, and then down to 1920x600px
• according to the window decoration, it's always maximized, so you have to click twice on the "maximize" icon to go to full screen after that.

Here are some relevant logs:

[2022-12-15T23:45:31Z WARN  widgetry::runner] Monitor scale factor is 1, screen window is 800, but the application requires 1500. Overriding the scale factor to 1.
[2022-12-15T23:45:31Z TRACE widgetry::runner] winit event says the window was resized from 800, 600 to ScreenDims { width: 1920.0, height: 1013.0 }. But inner size is ScreenDims { width: 1920.0, height: 1013.0 }, so using that
[2022-12-15T23:45:33Z TRACE widgetry::runner] winit event says the window was resized from 1920, 1013 to ScreenDims { width: 1920.0, height: 600.0 }. But inner size is ScreenDims { width: 1920.0, height: 600.0 }, so using that

I'm not sure why the initial window is 800x600px. From what I understand, the issue likely lies in glutin, but that crate was completely rewritten recently, so if there's a bug there it's not going to get fixed.

It looks like updating glutin isn't a straightforward task.

Madison designed an interactive tutorial mode for the LTN tool, but I haven't found time to implement it yet. If someone's not afraid of the slightly odd UI library A/B Street uses, it could be a fun medium-sized starter project. The design can be found at https://www.figma.com/file/HxSNM2146bcb0P0bNlW0x1/LTN-Working-Station-V2?node-id=5369%3A5464&t=cE5ZYUOuSsUMKwl2-0. Some samples: Embedding a second view of the map canvas is a bit advanced, but something we've figured out before. If anyone's interested in implementing, I can help flesh out smaller tasks and figure out how to structure the code.

CC @matthieu-foucault as an option I forgot from the other list

Goal: Adapting 15-min Santa game to an Easter Bunny version

Target users: AI UK attendees & Nick M's kids 😄

Needs: art/design, development support

Ideas brainstormed with @dabreegster:

• Switch Santa with Easter Bunny
• 2 player game: one is focused on dropping eggs and the other on finding eggs?
• Adapt to other cities across the UK and worldwide