'Crossing ways' shouldn't include grade separations

[NE2]

The 'crossing ways' validation would be a lot more useful if it didn't list cases where one of the following is true:
*Only one of the ways is a bridge or tunnel, or one is a bridge and the other a tunnel
*Both ways are bridges or tunnels, and the layer values are different
In other words, it would still list cases of bridges or tunnels crossing where the layer values are the same or one or both are not set.

[skyper]

Seems to be fixed.

Validator does not warn anymore, if the layers are different.

r4137

[NE2]

This wasn't fixed. Load ways 114877377 and 114877354 and run the validator. (JOSM should not enforce optional tags like layers on simple bridges.)

[skyper]

Replying to NE2:

This wasn't fixed. Load ways 114877377 and 114877354 and run the validator. (JOSM should not enforce optional tags like layers on simple bridges.)

So far, you have to set a layer tag to one of the ways, otherwise it is an error in the data. There was no resolution to ommit layer tags for bridges and tunnel and they do not imply it neither. In a matter of facts there was a vote on the wiki about it some month ago and it was rejected.

I think validator is right about the warning.

[NE2]

This isn't an error in the data. If you have a simple bridge, there's no ambiguity over what's above what.

[mrwojo]

Ticket #7655 has been marked as a duplicate of this ticket.

[anonymous]

Ticket #8111 has been marked as a duplicate of this ticket.

[skyper]

As far as I know it did the discussions about layer for tunnel and bridge not end up with an agreement to tag them without layer.

You can argue that it is not needed for simple cases but then you end up with more rules and special cases. I'd say the layer tag does not harm.

Yes, I have seen bridges in tunnels.

---

Updating/Extending the wiki is always welcome and the validator pages need definitly some extentions. Would be nice if we would have all messages explained with an description how to fix it.

[grinapo]

As I have mentioned in the duplicate ticket people tend to mix up separate questions.

1) layer=* tag is useful, required, generally. No question.

2) there are constructions where layer=* tagging is required on bridges, tunnels, bridges on bridges, multilayered constructs. there is no ambiguity here, no question.

3) the question are simple bridges (bridge over one or a few ways, neither of them having layer=* tag) and simple tunnels (same way all unlayered). the "vote" was rejected, as far as I see because the vote email was held in spam queue. there was no resolution, no decision whether layer is required or superfluous for bridge=*.

The problem is that approximately 30% of the total bridge tags are not layer tagged so this is not a minority problem.

The mappers are held in ambiguity, some do this and others do that. Either the validator should support the defaults or it should warn about bridges/tunnels without layer tag, I believe, but I am no way authoritative here.

[simon04]

In 6579/josm:

fix #4280 - Validator, crossing ways: do not warn if layer is missing for simple, unambiguous cases (one way is bridge/tunnel and the other one is not)

[aceman]

Not sure why JOSM developer suddenly implemented this when other one was against it.

Anyway, what happens if there is a bridge without layer over some ways without layer. And then another bridge over all those and has a layer=1 set? Is there a warning for this ?

[skyper]

Replying to aceman:

Not sure why JOSM developer suddenly implemented this when other one was against it.

Me neither.

Anyway, what happens if there is a bridge without layer over some ways without layer. And then another bridge over all those and has a layer=1 set? Is there a warning for this ?

+1

As far as I know there was no decision on tagging@osm and the status quo is still to always add a layer tag.

[simon04]

In 6635/josm:

fix #4280 - Validator, crossing ways: undo valdation relaxation (one way is bridge/tunnel and the other one is not) - reverts r6579