TN: KY-1076

[rickmastfan67]

Yes, it seems that a small segment is officially in TN.  I was looking at this area do get the US-127 boarder problem fixed there for the TN side.
https://goo.gl/maps/nVTywrJEbmz

Do we just want to fudge KY-1076 to intersect US-127, or do we want to create a separate file?

Either way, we would have to update the 'TN/KY' point because of the reroute in that area.
http://tm.teresco.org/devel/hb.php?r=ky.ky1076

[yakra]

carrying over some discussion from the old forum:
http://clinched.s2.bizhat.com/viewtopic.php?t=2088

The 'old' location for TN/KY isn't part of KY-1076 anymore, as that intersection was moved a little bit to the East (see updated Google Sat Imagery as well as OSM). Since I couldn't ID any street name at the old state border, I've relabeled it at 'OldTN/KY', but I'm open for any other suggestions till it's pulled in (GitHub) and can adjust it.

I think a reference to a road name, even if just OldUS127_Sta here, is more appropriate. More intuitive than OldTN/KY IMO, and there's no precedent AFAIK for Old-[border] labels.

Also, since it seems that the TN/KY border is just a tick North of the intersection with TN-111, the TN/KY location in KY's file will need to be updated as well to reflect this since it seems to be possible to turn left onto TN-111 without going into KY.

Mm, I kinda wanna handle that differently. Judging by OpenMQ (or Bing), the stateline is right on the road centerline. I'm inclined to treat the road as the state line and vice versa, just like US59/71 in Texarkana, or ME NH153.
So, just one TN/KY point, where TN111Sta is now, which could still be used as a waypoint for those who turned onto TN111. (The TN/KY point in the KY file would be positioned here too.)
(TN111Sta, if kept, would need to be relabeled -- either as TN111_Sta, or more properly, TN111_N, with a conforming change to vanilla TN111 farther south.)

Still, this leaves us with the KY1076 issue...

Yes, it seems that a small segment is officially in TN.

You do see the county line sign right there in GMSV. But any way you slice it, whichever map flavor we're looking at and wherever it puts the state line, I've got to agree. Small tiny segment officially in TN.

Do we just want to fudge KY-1076 to intersect US-127, or do we want to create a separate file?

I'm inclined to fudge it one way or another and only have the one file, already created. Do we...
* Fudge it to end at the state line, and not quite US-127? Or...
* Fudge it to intersect US127, a little past the state line?
I regard the former as the lesser evil. With the caveat that I wonder how it would affect Jim's graphing algorithms. I also idly wonder how it would affect CHM's algorithms to pick up on intersecting routes. Unfortunately there's no way to get it into CHM to test that out. Jim, how do your graph algorithms handle NMPs? Would points, say, here and here qualify as NMPs?

Either way, we would have to update the 'TN/KY' point because of the reroute in that area.
http://tm.teresco.org/devel/hb.php?r=ky.ky1076

Correct.
And it looks like I missed this subtlety in my last post on the old forum, when suggesting how to relabel existing TN/KY on US127.

[rickmastfan67]

carrying over some discussion from the old forum:
http://clinched.s2.bizhat.com/viewtopic.php?t=2088

The 'old' location for TN/KY isn't part of KY-1076 anymore, as that intersection was moved a little bit to the East (see updated Google Sat Imagery as well as OSM). Since I couldn't ID any street name at the old state border, I've relabeled it at 'OldTN/KY', but I'm open for any other suggestions till it's pulled in (GitHub) and can adjust it.

I think a reference to a road name, even if just OldUS127_Sta here, is more appropriate. More intuitive than OldTN/KY IMO, and there's no precedent AFAIK for Old-[border] labels.

Well, I mentioned this in the pull request, and Jim didn't seem to have any problems with it.  But, I could change it.

Also, since it seems that the TN/KY border is just a tick North of the intersection with TN-111, the TN/KY location in KY's file will need to be updated as well to reflect this since it seems to be possible to turn left onto TN-111 without going into KY.

Mm, I kinda wanna handle that differently. Judging by OpenMQ (or Bing), the stateline is right on the road centerline. I'm inclined to treat the road as the state line and vice versa, just like US59/71 in Texarkana, or ME NH153.
So, just one TN/KY point, where TN111Sta is now, which could still be used as a waypoint for those who turned onto TN111. (The TN/KY point in the KY file would be positioned here too.)
(TN111Sta, if kept, would need to be relabeled -- either as TN111_Sta, or more properly, TN111_N, with a conforming change to vanilla TN111 farther south.)

Well, looking @ both Google & OSM proper, they both show the border clearly North of the new alignment of US-127.  Do we know of any official KY or TN GIS source that we can verify the border there?  Google looks to be the correct one, but who knows.

Still, this leaves us with the KY1076 issue...

Yes, it seems that a small segment is officially in TN.

You do see the county line sign right there in GMSV. But any way you slice it, whichever map flavor we're looking at and wherever it puts the state line, I've got to agree. Small tiny segment officially in TN.

Do we just want to fudge KY-1076 to intersect US-127, or do we want to create a separate file?

I'm inclined to fudge it one way or another and only have the one file, already created. Do we...
* Fudge it to end at the state line, and not quite US-127? Or...
* Fudge it to intersect US127, a little past the state line?
I regard the former as the lesser evil. With the caveat that I wonder how it would affect Jim's graphing algorithms. I also idly wonder how it would affect CHM's algorithms to pick up on intersecting routes. Unfortunately there's no way to get it into CHM to test that out. Jim, how do your graph algorithms handle NMPs? Would points, say, here and here qualify as NMPs?

Either way, we would have to update the 'TN/KY' point because of the reroute in that area.
http://tm.teresco.org/devel/hb.php?r=ky.ky1076

Correct.
And it looks like I missed this subtlety in my last post on the old forum, when suggesting how to relabel existing TN/KY on US127.

Well, we do have very short segments in other states.  0.1 might just qualify if Google's location of the state border for TN/KY here is correct.

[yakra]

Well, I mentioned this in the pull request, and Jim didn't seem to have any problems with it.  But, I could change it.

Yes, let's change it.

Well, looking @ both Google & OSM proper, they both show the border clearly North of the new alignment of US-127.  Do we know of any official KY or TN GIS source that we can verify the border there?  Google looks to be the correct one, but who knows.

I searched around a bit; no dice on GIS for either state. The KYTC SPRS maps don't show the state line segment of US127, so that supports putting it in the TN file. Thinking about it a bit, and thinking back on the NH ME113 / NH113B situation, I'm fine with keeping the route trace as you have it in TN, with the road slightly south of the state line, and separate points for TN111 and the state line.
I still recommend label changes to the TN111 points, however:
TN111 -> TN111_S
TN111Sta -> TN111_N

Well, we do have very short segments in other states.  0.1 might just qualify if Google's location of the state border for TN/KY here is correct.

If you're in favor of having two route files for KY1076, I can be on board with that too. Let me know, and I'll make a 2nd route file and adjust the CSVs accordingly.

[rickmastfan67]

yakra, could you submit the label fix for the old border location (and a fix for the 'update' log's old entry)? Still don't have my pc back online yet.

As for the TN-111 labels, I don't recommend the changes you suggested.  I would go with TN111_Sta instead for the northern one since the current 'TN111' point is an intersection, and not a start/end of a multiplex, thus not totally justifing the _S & _N parts.

As for a TN segment for KY-1076, yes, I think that would be the best thing to do unless we ever do get 100% proof of the exact location for the state border there.

[yakra]

yakra, could you submit the label fix for the old border location (and a fix for the 'update' log's old entry)? Still don't have my pc back online yet.

Done. Also, OldUS127_S & _N -> OldUS127_JamS & _JamN.

As for the TN-111 labels, I don't recommend the changes you suggested.  I would go with TN111_Sta instead for the northern one since the current 'TN111' point is an intersection, and not a start/end of a multiplex, thus not totally justifing the _S & _N parts.

Hehee... And the points aren't the end of a child route either. For these reasons, I've also historically preferred city suffixes on labels, though I eventually came round to the canonical form: _S & _N since they're the southern & northern junctions with this route, along the route being mapped.
I have no problem with it though, and there's enough precedent. Changed to TN111_Dun and TN111_Sta.

As for a TN segment for KY-1076, yes, I think that would be the best thing to do unless we ever do get 100% proof of the exact location for the state border there.

If we *do* get 100% proof of the border, I bet it'd just confirm what we already have, that the roadway is south of the border at this location. TN KY1076 added.

https://github.com/TravelMapping/HighwayData/pull/364
US127 diff

[yakra]

OK to merge it in?

[rickmastfan67]

You saw my comment over on GitHub.  I think we can mark this as 'solved' now.