NY: NY14A/54 concurrency broken

[Duke87]

NY 54 has a NY14A_S point but no corresponding NY 14A_N point.

Based on signs the northbound routing of NY 54 through Penn Yan is Liberty-Elm-Main, but the southbound routing is Main-Water-Liberty. So, the NY14A_N point should be halfway between Elm and Water, though it'd be tempting to fudge it to Elm to get a graph connection with NY 54A.

[yakra]

Edited NY54 a few years back.
I thought there was a prior topic on this, but maybe not; nothing was linked from the pull request.
NY54 cuts straight across the couplet without favoring one direction or the other.
There's a graph connection at NY14A_S/NY54_S.
No-build.

[neroute2]

I'm confused. Northbound NY 54 traffic is directed to turn left on NY 14A. Southbound NY 54 traffic uses Water Street. So no matter which way you go on NY 54, you're using NY 14A south of Water Street. That's very clearly an overlap that needs to be mapped.
(I would also put a point on NY 54 at the Elm/Main intersection to make the routing clearer and potentially avoid questions, since OSM doesn't show it properly.)

[Duke87]

Edited NY54 a few years back.

So... it looks like it was correct before it was changed.

What's in the HB for NY 54 currently would be fine if one of the directions used Lake east of 14A, but neither does.

(I would also put a point on NY 54 at the Elm/Main intersection to make the routing clearer and potentially avoid questions, since OSM doesn't show it properly.)

I agree with this and think it's best practice to have a point at any intersection that's a divergence point for different directions of a route regardless.

[yakra]

I'm confused. Northbound NY 54 traffic is directed to turn left on NY 14A. Southbound NY 54 traffic uses Water Street. So no matter which way you go on NY 54, you're using NY 14A south of Water Street. That's very clearly an overlap that needs to be mapped.
(I would also put a point on NY 54 at the Elm/Main intersection to make the routing clearer and potentially avoid questions, since OSM doesn't show it properly.)

Good point. Done.
https://github.com/TravelMapping/HighwayData/pull/6861

So... it looks like it was correct before it was changed.

It would appear so. In location if arguably not in label. Maybe that was what threw me off/ brought it to my attention. *Shrug.*

I agree with this and think it's best practice to have a point at any intersection that's a divergence point for different directions of a route regardless.

I don't necessarily think they're always needed bar none; there'll be insignificant cases where it really doesn't matter.
That said, I do agree with neroute2 and think a point is useful for clarification here. Added.