tilezen / vector-datasource

Tilezen vector tile service - OpenStreetMap data in several formats
https://www.nextzen.org/
Other
507 stars 119 forks source link

Add ref values for railway=station_entrances and aeroway=gate #527

Closed nvkelso closed 8 years ago

nvkelso commented 8 years ago

See also:

Based on feedback from Korea test here:

https://www.openstreetmap.org/edit?node=3824606390#map=19/37.50438/127.02449

We'd expose this in map display styles by concatenating the ref and name (like we do roads today). Once shields become a Tangram feature, we'd transition to a hybrid solution.

nvkelso commented 8 years ago

This looks to in prod, at least for subway entrances:

19/48.86814/2.36145

screen shot 2016-02-16 at 17 29 36

zerebubuth commented 8 years ago

I'm not sure I understand - what do we have to add that's not already in the tiles?

From your screenshot, it looks like we've already got ref for kind: subway_entrance, and from the screenshot below you can see we've got ref for kind: gate, aeroway: gate:

aeroway-gate

The actual scene file snippet is:

    aeroway-gate:
        data: { source: osm, layer: [pois] }
        filter: { kind: gate, aeroway: gate }
        draw:
            icons:
                visible: true
                interactive: true
            text:
                visible: true
                interactive: true
                text_source: ref
                font:
                    fill: '#ff0000'
                    size: [[12,9px],[15,12px],[17,14px]]

But I wasn't able to get the interactive bit to work.

From taginfo, it looks like there are 7 railway=station_entrance objects, and 131 railway=train_station_entrance objects. It's likely that we'd have better results looking for entrance=* on the outline of the station polygon, or building polygons within the station site relation (there are 991,072 entrance=* nodes, but it's not easy to tell how many are part of polygons). But that's a lot more work.

nvkelso commented 8 years ago

Agree there's nothing to do here – the issue in Korea is the data hasn't been tagged properly yet. Once it is tagged in full detail, we'd pass that thru.

Let's defer the entrance=* idea for later if it comes up again.