[Tagging] Feature Proposal - RFC - Railway tracks on highway

Nikulainen, Jukka K jukka.nikulainen at helsinki.fi
Fri Dec 14 08:00:52 UTC 2018

Hello all!

And sorry for the delay in my comments, I've been very busy with other work the past week.

Firstly, I was rethinking about the introduction a general "=yes"-value for this tag. It might be that some lazy mappers would prefer this kind of general tag, but on the other hand the "embedded_rails=*"-tag perhaps should always simply inherit its value directly from the "railway=*"-tag it shadows, so figuring out the proper specific tag should not be a problem. On the other hand, a "=no"-value could be useful in combination with the ":lanes=" subkey to explicitly mark the absence of an embedded rail on a specific lane. The :lanes wiki page does not require such an explicit "no"-value, but perhaps this would nonetheless be useful in this circumstance.

Perhaps the proposal should be changed to make it explicit that the "embedded_rails=*" value should always inherit the value of the "railway=*" tag it shadows, regardless of what that value is?

Paul also had a very good point about the problems associated with the "=abandoned" and "=disused"-values. It's certainly true that the lifecycle subtags should be used here, but also true that the disused railway lanes still, physically, exist in the ground, even if disused. Furthermore a quick search on overpass turbo shows that the "railway=disused" tag is used all over the place! The overpass web frontend could't even process the whole of central Europe at once because of the frequency of the tag. Also, with regards to the point I made earlier, it could make sense to have the "embedded_rails=*" tag always shadow the "railway=*" value. Then at least the use would be consistent, even if consistently "wrong" :).


More information about the Tagging mailing list