CT: CT 15, US 5, CT 2 around Hartford

[vdeane]

I was browsing around this area today, and the points at these interchanges seem a little odd.  US 5 and CT 15 have the point where US 5 splits off centered where the ramps enter/exit rather than on Main Street, seemingly trying to capture East River Drive even though only one direction of US 5 would use it.  Neither US 5 nor CT 15 has a point at CT 2, and CT 2's 4A point is located a little south of US 5/CT 15.

[yakra]

Kind of an icky area here, with no real perfect way to handle it.

US 5 and CT 15 have the point where US 5 splits off centered where the ramps enter/exit rather than on Main Street, seemingly trying to capture East River Drive even though only one direction of US 5 would use it.

Exactly. Treating this as a "misbehaving parclo" and cutting across the diagonal of the (NB ramps & Main / E River & SB ramps) couplet, between where the NB & SB alignments split at the ramp gores, to where they rejoin at SilLn.

Neither US 5 nor CT 15 has a point at CT 2,

In this case, the ramp leaves the mainline along with the Main St / US5 N ramp, so I opted to go "1 point per interchange" and fold it into the 5/15 split point.

and CT 2's 4A point is located a little south of US 5/CT 15.

This was informed by the odd, onramps-only layout, connecting here.

[Duke87]

One drawback of the implementation as it currently exists is that CT 2 does not have a graph connection with US 5 or CT 15.

[yakra]

True. Thought of that, waited for someone to mention it. And you did. I'm amenable to adding a graph connection here.

[yakra]

Decided against doing this. An extra point for an exit on CT15 that already has a waypoint, which would result in ugly out-of-order alphabet soup, and bunch together a couple points on CT2 that are fairly evenly spaced right now. Ick, no thanks.
• 1PPI
• points @ connector roads like trumpets and the like.