PA: PA 372 Point in Buck (near PA 272)

[Markkos1992]

The HolRd_E (where PA 372 turns south to head to PA 272) point needs to be recentered onto PA 372.  Though PA 372's routing in general is confusing here. 

[yakra]

Another fun one to look at in HistoricAerials. The HolRd_E point (which should be labeled OldHolRd?) is placed about where Four Pines Drive tied in before the bypass bas constructed. Per OSM, PA372 (in just one direction? Both?) goes around to the east on Friendly Drive as well. Indeed, the left turn movement at PA272_N (from Buck Road onto PA272 south) is not possible; PA272 is divided there.
Looks as if this might have been built to have the 272/372 intersection on a flatter grade.
I'll look into it a little bit more before making any edits just to make sure I have everything right.

[yakra]

Right. County maps are too lo-res to get any meaningful info. Per township maps and the MPMS-IQ utility, the southern bypass is SR 3372. SR 0372 is still on Holtwood (apparently not "Old" Holtwood as OSM would suggest) and Buck Roads, straight as the crow flies but no left turns or cross access at PA272.
AIUI, there's not always a 1:1 correspondence between the 4-digit SR numbers and PA routes with the keystone shield. Right? (Markkos1992, did I see you say you work for PENNDOT? Any insight here?) Certainly, PA372 isn't signed on Holtwood Rd between HolRd_E and PA272.

That all said, on to GMSV...
The eastern part of the bypass, from HolRd_E, is signed for PA372 in both directions. As shown in the HB.
No imagery on the south end of the bypass facing PA272.
Facing north on PA272, signage suggests a routing as shown in the HB. Here too.
I don't see any PA372 shields on Friendly Dr itself. (This is too blurry to count )
Here's the kicker: West on PA372 at Friendly Dr. TO WEST PA372. A pretty solid indication it's not considered part of PA372 itself. However, there's no other way to make the movement to continue on PA372.
So we've got a bit of an *ahem* "operational" couplet: NB as shown in the HB via PA272 and Buck Rd, and SB via Friendly Dr.

As such, I'm a bit inclined to cut the route trace across the middle of the "couplet", not favoring any particular direction, and replace PA272_N with a point at Buck Rd and Friendly Dr. Thoughts?

In any case, recentering HolRd_E will require editing the Sharp Angle datacheck entry for PA272_S.

[Markkos1992]

No true insight as I work for Central Office and not District 8.  Also note the 1:1 thing is technically correct because US 30 Business (a 4-digit SR Number I do not remember offhand) for instance cannot be SR 30 as US 30 is SR 30.  Having said that, the VideoLog map agrees with your findings. 

I agree with the Buck Rd at Friendly Dr point though I think the PA 272 point itself should stay as-is. 

[rickmastfan67]

AIUI, there's not always a 1:1 correspondence between the 4-digit SR numbers and PA routes with the keystone shield. Right? (Markkos1992, did I see you say you work for PENNDOT? Any insight here?) Certainly, PA372 isn't signed on Holtwood Rd between HolRd_E and PA272.

Yep, sometimes PA (keystone) routes can have a different 'SR' number.  PA-380 is SR-0400 (because SR-0380 is used for I-380), PA-99 has SR-0699 (though used to have SR-0099) because I-99 has SR-0099.  There are also a few other examples out there.

==

Anyways, I would say just keep the PA-272 points as-is and just recenter HolRd_E & add the point @ Friendly Drive.  Probably the best solution here.

[yakra]

Anyways, I would say just keep the PA-272 points as-is and just recenter HolRd_E & add the point @ Friendly Drive.  Probably the best solution here.

I still can't fully bring myself to keeping the PA272_N point -- such a routing cannot be traveled in the westbound direction.
If it makes a difference,  the change can be made without breaking any .list files:
• imgoph has only the PA272 concurrency marked, and it's also marked on PA272 itself. No problem.
• dave1693's travels will still be parsed correctly if we have a PA272 +PA272_S +PA272_N (Or actually, just PA272 +PA272_N) point.

It's kinda like... an implied concurrency. On an unsigned route. On just one direction of a couplet...

[yakra]

HolRd_E recentered; FriDr added; PA272_N left as-is.
I still don't really agree with that, but it seems to be the consensus, and is the smaller change from status quo. I just don't really disagree enough to care much more.
Still open to cutting the diagonal across the couplet in the future as appropriate.
I've asked a local non-roadgeek friend for photos of these signs, but nonetheless:
https://github.com/TravelMapping/HighwayData/pull/1665

In any case, I'll leave this topic marked unsolved at least until the datacheck fix is resolved.
Edit:/ https://github.com/TravelMapping/HighwayData/pull/1670