Graphhopper VS OpenLayers3

Well I think what he means is that we dropped the explicit support of Android and the example app with offline maps (Mapsforge), see this issue: https://github.com/graphhopper/graphhopper/issues/1940

But offline routing should still work on Android. But as you need JDK 1.8 support this will exclude a few older Android versions and devices, I think.

Graphhopper. It has an Android client that uses its web service. You can also try out the web app.

This is now fixed and deployed. See the route here and the specific fixed examples here. Note that you can still find the dangerous spots via clicking the "stop" icon below the route information multiple times. But this is likely a result of the bad cycling infrastruture as cycle.travel contains these short dangerous parts too and if you do such longer routes for Germany it does not happen.

I'd suggest refactoring to use a self hosted version of Graphhopper (https://www.graphhopper.com/open-source/ - they have a docker image you can use and pass in an OSM file to build a graph on) or OpenRouteService to limit your cost exposure to google.

Open Source. The routing server is open source too and geocoding too.

Valhalla is great (I've contributed a few patches and run it for a while)!

Also check:

http://project-osrm.org/

https://www.graphhopper.com/open-source/

if you attached a geolocation to each meeting that should be possible. either using the google maps API or a self-hosted graphHopper instance

I just improved the documentation for the map matching a bit. Maybe this also helps a bit.

You probably know this, but in Google Maps at least, you can use browser zoom (ctrl/cmd +/-) to change the size of labels without zooming into the actual map.

------

Speaking of maps, I got to work a fun zoom project a few years ago: https://map.fieldmuseum.org/

We used https://openlayers.org/ and thought long and hard about how to best handle zooming and variable levels of information density & visual hierarchy. If you zoom all the way out, we just highlight where the building is relative to the surroundings. As you start to zoom in, we start to highlight major exhibitions and entrances. Then as you zoom in more, we start showing recommended paths, smaller exhibitions, etc. The label sizes try to scale up and down at each level, smoothly, in order to balance readability and density.

Eventually you can reach the max zoom level and the labels will just grow bigger and bigger, but the SVGs dynamically shrink so they remain pictograms and not just contextless-lines.

Then if you keep going, you eventually find microscopic easter eggs :)

The code is pretty jank (and abandoned), but it's FOSS vanilla JS/HTML/CSS, and the only dependency is on OpenLayers: https://github.com/arcataroger/openlayers_indoor_map

In order to display the GeoJSON features on a map, we will use OpenLayers, which is a very powerful open-source mapping library that is also very simple to use.

OpenLayers is available via the ol npm package, offering developers a powerful toolkit for creating sophisticated maps. Here is a JavaScript implementation that utilizes OpenLayers to showcase a map:

Official Website: https://openlayers.org/

For web maps I'd strongly recommend using OpenLayers. While it's less convenient to get started with compared to the alternatives it's also much more feature-complete and you'll likely hit a ceiling in terms of functionality much later than you would with the others.