NJ: I-295/US-30 interchange

[rickmastfan67]

Shouldn't we have a graph connection here between the two routes?  There are 3 direct ramps between each of them that would warrant a point centered on the overpass there.

What I would change is the following:
I-295:
NEW -> 29A (centered on the overpass of I-295 over US-30)
29 -> 29B +29

US-30:
NEW -> I-295 (centered on the overpass of I-295 over US-30)
I-295 (current) -> CR666

While the 'I-295' point would have to move, it's close enough that I'd think no update mention for the update page would be necessary in this case.

[yakra]

I can see maybe changing US30... the I-295 crossing is just about exactly centrally located in the interchange footprint on that road. And it's more intuitive to have an I-295 point there, rather than at a separate at-grade jct that would otherwise be CopRd.

I-295?
29B wouldn't be our label for Copley Rd, as the 29B ramp connects only to US30, same as 29A.
But regardless... Extra points?

Is this complex a case of...
• a single interchange
• double half interchanges, or
• two interchanges?
The way the ramps of the one half fill out the missing moves of the other, it falls more into the "one interchange" camp than two. IMO it's a "misbehaving diamond" variant centered at Copley Rd with some extra/redundant moves thrown in.

If we consider it to be double half interchanges, it doesn't pass the "clear gap of at least 0.5 mi/0.8 km separates the two halves, or each half connects to a different highway that we are also mapping" test. Ergo it's a 1PPI situation; "use one central point and treat both halves as a single, full interchange." Falls into the exceptions to positioning at centerline crossings and unusual shapes categories, IMO.
One central point midway between the westernmost & easternmost gores neatly corresponds to midway between the two centerlines we're serving, so w0ot. And it's a good location to indicate that the point is intended to serve the whole interchange complex, rather than just half of it.

While I do see some room for improvement, sadly we still don't quite make it to a graph connection here.
https://github.com/TravelMapping/HighwayData/pull/4196

[rickmastfan67]

I would still say that we should have the graph connection due to both routes being ones we map, and there are direct ramps between the two, just saying.  It's not like the ramps are a mile long access road to a trumpet where it's impossible to have a graph connection.