search

abrensch / brouter

configurable OSM offline router with elevation awareness, Java + Android
MIT License
510 stars 122 forks source link

Please consider cycleway:surface in lookups.dat #131

Closed utack closed 3 years ago

utack commented 5 years ago

The tagging for cycleway surface using "cycleway:surface" is on the rise, according to the OSM weekly blog. It is used on over 5300 ways already. https://wiki.openstreetmap.org/wiki/Key:cycleway:surface I think this is a great key to have for a bike router, please consider adding it to lookups.dat Thank you

export

Phyks commented 5 years ago

Hi!

This is done in this commit (not yet part of a PR). Could you have a look and let me know what you think about it? I would be interesting about second opinions about the profiles part in particular :)

Best,

Phyks commented 5 years ago

Hi @utack, would you have a specific example in mind where taking cycleway surface into account would change the result from the routing? I'd like to add a unittest for it in https://github.com/phyks/broutertesting. Thanks!

poutnikl commented 5 years ago

Why should have cycleway:surface= different routing effect, compared to surface= ?

Physical effect of surface on biker experience and preference does not depend on how it is mapped.

If a cycleway has an unpleasant surface, no matter how or if the surface is tagged, a biker may prefer other road, if traffic rules allow it.

Phyks commented 5 years ago

Typically, I can imagine a paved street having a cycleway with asphalt surface. Then, with the current profiles, the cycleway would be considered as having the surface (which is the one of the street) and be penalized when in fact the cycleway:surface would leverage this penalty.

utack commented 5 years ago

@Phyks I will get back to you
The real-world cases I can think of are vast (cobblestone on the bike path, asphalt on the road), but the mapped ones are still rare. The ones I found I will report to you!

poutnikl commented 5 years ago

@Phyks I do agree with you. If cycleway:surface= has significant occurence, it should be evaluated with the same weight but higher priority than surface=

Phyks commented 5 years ago

@utack Thanks! You can also add some in OSM data if you know them and they are not mapped :p

@poutnikl Exactly, that's what I'd like to add a test in https://github.com/Phyks/BrouterTesting for. :)

theraser commented 3 years ago

I just noticed that the changes at lookups.dat are already applied, maybe we can close this issue here? Or is there something more to do?

Sadly the cycleway:surface tags aren't used by brouter-web, I just created an issue over there: https://github.com/nrenner/brouter-web/issues/438