Points on concurrent routes mixed hidden/visible?

[Jim]

Can anyone think of a reason we'd want to allow a waypoint that's somewhere along a concurrency of two or more routes where that waypoint's primary label is hidden on one route but visible on another?  I'm not talking about hidden alternate labels, which are ignored in the situation I'm considering.  I'm making the assumption that a point is either hidden or visible, consistently among all routes that might share that point.

[michih]

I agree. A location should not be hidden on a route and visible on a concurrent one. If it's hidden, it's an error.

[Jim]

Related to this, I was also assuming that no hidden waypoint could be at a junction with another TM highway.  Turns out this is not always the case.  For example, we used to have an erroneous "intersection" (which I define for my purposes as two routes crossing and both have a waypoint at precisely the same coordinates) between NY 5S and NY 171.  That was removed from our system by hiding the old waypoints (NY5S in ny.ny171 and NY171 in ny.ny005s) but the points remained at idential coordinates.  I'm going to "fix" this situation by moving NY171's point to a nearby intersection and unhiding.  I'll probably do similar things in other cases I find.  If I think there's any potential problem from one of these changes, I'll post a note here for comment.

[Jim]

OK, so I already have another example and this one might not be so simple.  US 1 in CT at its intersection with US 7 has 4(!) waypoints, which to me seems to violate the one point per interchange rule.  The immediate concern to me is the hidden US7 point in the middle of the mess, concurrent with US 7's US1 point.  It is not in use in TM, so I could just remove it, but perhaps we can get a few people to take a look at that area and decide if it makes more sense to have single US7 point, exactly where +US7 is now, and get rid of US7_S, US7_N, and RivAve (keeping US7_S as a hidden alternate for US7, as that point is in use).

[Jim]

There are currently 50 instances I detected of the situation I described in the OP.  Places where the same coordinates are used on two routes, either intersecting or as part of a concurrency, but one route's waypoint is hidden while the other is visible.  Thoughts?

Code: [Select]

ERROR: hidden waypoint colocated with visible, sct.e5gla +M74 (55.851204,-4.272523) and sct.e5 M8/M74 (55.851204,-4.272523)
ERROR: hidden waypoint colocated with visible, sct.e16 +M74 (55.851204,-4.272523) and sct.e5 M8/M74 (55.851204,-4.272523)
ERROR: hidden waypoint colocated with visible, bel.e19 +N133 (51.35364,4.630265) and bel.a001 *N133 (51.35364,4.630265)
ERROR: hidden waypoint colocated with visible, che.e27 +11(A16) (47.351545,7.366934) and che.a016 11 (47.351545,7.366934)
ERROR: hidden waypoint colocated with visible, swe.e45 Avv (67.330253,21.078579) and swe.e10 +X34 (67.330253,21.078579)
ERROR: hidden waypoint colocated with visible, cze.e55 +52(D8) (50.527479,13.999629) and cze.d008 52 (50.527479,13.999629)
ERROR: hidden waypoint colocated with visible, aut.e55 +A7/S10 (48.356257,14.460582) and aut.a007 S10 [alt: ['+B125']] (48.356257,14.460582)
ERROR: hidden waypoint colocated with visible, aut.e461 +S1/S2 (48.285863,16.514103) and aut.s001wie 36 [alt: ['+B8']] (48.285863,16.514103)
ERROR: hidden waypoint colocated with visible, esp.a001vit NC/PV (42.93272,-2.227993) and esp.e5bea +NC/PV (42.93272,-2.227993)
ERROR: hidden waypoint colocated with visible, esp.b030 4 (41.491912,2.095385) and esp.e15 +4(B30) (41.491912,2.095385)
ERROR: hidden waypoint colocated with visible, fra.a006a +N186 (48.756189,2.334294) and fra.e15 N186 [alt: ['+A86']] (48.756189,2.334294)
ERROR: hidden waypoint colocated with visible, sct.m008 +M74 (55.851204,-4.272523) and sct.e5 M8/M74 (55.851204,-4.272523)
ERROR: hidden waypoint colocated with visible, sct.m074 +M8 (55.851204,-4.272523) and sct.e5 M8/M74 (55.851204,-4.272523)
ERROR: hidden waypoint colocated with visible, eng.a001m +X27 (54.261565,-1.517572) and eng.e15 *B6285 (54.261565,-1.517572)
ERROR: hidden waypoint colocated with visible, eng.a001m +X28 (54.300633,-1.567305) and eng.e15 *A684 (54.300633,-1.567305)
ERROR: hidden waypoint colocated with visible, eng.a5080 M62(4) (53.405995,-2.89624) and eng.e20liv +M62(4) (53.405995,-2.89624)
ERROR: hidden waypoint colocated with visible, ita.ra001 +ViaMic (44.519023,11.371815) and ita.e45 ViaMic (44.519023,11.371815)
ERROR: hidden waypoint colocated with visible, pol.a002min 541 (52.197875,21.775675) and pol.e30 +541(A2) (52.197875,21.775675)
ERROR: hidden waypoint colocated with visible, pol.a004 581 (50.093564,22.058895) and pol.e40 +581(A4) (50.093564,22.058895)
ERROR: hidden waypoint colocated with visible, pol.s011 +270 (52.345722,16.754751) and pol.s005poz 201 (52.345722,16.754751)
ERROR: hidden waypoint colocated with visible, svk.d001 187 [alt: ['+D3']] (49.225593,18.608887) and svk.e50 +D1/D3 (49.225593,18.608887)
ERROR: hidden waypoint colocated with visible, svk.d003 0 [alt: ['D1']] (49.225593,18.608887) and svk.e50 +D1/D3 (49.225593,18.608887)
ERROR: hidden waypoint colocated with visible, tur.o005 16 (40.28974,29.025192) and tur.e90 +16(O5) (40.28974,29.025192)
ERROR: hidden waypoint colocated with visible, tur.o020 +6 (39.786824,32.742562) and tur.e90 6(O20) (39.786824,32.742562)
ERROR: hidden waypoint colocated with visible, tur.o022 O5 (40.28974,29.025192) and tur.e90 +16(O5) (40.28974,29.025192)
ERROR: hidden waypoint colocated with visible, ct.us007 US1 (41.118111,-73.419893) and ct.us001 +US7 (41.118111,-73.419893)
ERROR: hidden waypoint colocated with visible, ms.us078 +Fut269 (34.87896,-89.724784) and ms.i022futtup Fut269 (34.87896,-89.724784)
ERROR: hidden waypoint colocated with visible, co.us087 I-25(137) [alt: ['+137(I-25)']] (38.78321,-104.788091) and co.i025 +137 (38.78321,-104.788091)
ERROR: hidden waypoint colocated with visible, me.me170 StaFacRd (45.421144,-68.137025) and me.me169 +x0StaFacRd (45.421144,-68.137025)
ERROR: hidden waypoint colocated with visible, nv.nv599 JonBlvd (36.231154,-115.224974) and nv.us095buslsv +NV596 (36.231154,-115.224974)
ERROR: hidden waypoint colocated with visible, oh.oh043 +x240 (41.461389,-81.642509) and oh.oh014 AetAve (41.461389,-81.642509)
ERROR: hidden waypoint colocated with visible, oh.oh066 +x170 (41.288203,-84.361152) and oh.oh015 FortSt (41.288203,-84.361152)
ERROR: hidden waypoint colocated with visible, oh.oh328 +y91 (39.399007,-82.372897) and oh.oh056 LauRunRd (39.399007,-82.372897)
ERROR: hidden waypoint colocated with visible, aze.ah081 +X187946 (39.869235,48.232641) and aze.e002 Mur (39.869235,48.232641)
ERROR: hidden waypoint colocated with visible, bc.bc001 +x4 (49.227183,-122.81996) and bc.tchmai 44A (49.227183,-122.81996)
ERROR: hidden waypoint colocated with visible, chnnx.g70 +G22_W (35.591017,106.306715) and chnnx.g22 G70_W (35.591017,106.306715)
ERROR: hidden waypoint colocated with visible, chngx.g72 +G65/G76 (25.280441,110.158281) and chngx.g65 G72_E (25.280441,110.158281)
ERROR: hidden waypoint colocated with visible, irn.f001 R49_Man (36.730974,49.418435) and irn.ah008 +F1 (36.730974,49.418435)
ERROR: hidden waypoint colocated with visible, kor.ex065 7 (35.553737,129.249344) and kor.ah006 +7(Ex65) (35.553737,129.249344)
ERROR: hidden waypoint colocated with visible, kor.ex065och 10 (35.783006,129.424438) and kor.ah006 +10(Ex65) (35.783006,129.424438)
ERROR: hidden waypoint colocated with visible, kor.ex065gan 31 (37.512245,129.095535) and kor.ah006 +31(Ex65) (37.512245,129.095535)
ERROR: hidden waypoint colocated with visible, svk.i002 PanCes_N (48.131805,17.104812) and svk.e58 +1A(D1) (48.131805,17.104812)
ERROR: hidden waypoint colocated with visible, in.in003 +X00 (39.802074,-85.44548) and in.us040 *IN3 (39.802074,-85.44548)
ERROR: hidden waypoint colocated with visible, idn.ah002 +266 (-6.898735,108.880091) and idn.jtkp 266 (-6.898735,108.880091)
ERROR: hidden waypoint colocated with visible, bgr.ii066 +A4 (42.181405,25.246625) and bgr.a004 *II66 [alt: ['+N66']] (42.181405,25.246625)
ERROR: hidden waypoint colocated with visible, nl.nl430 +Goo/Yah (51.31232,-56.708475) and nl.nl430 MisCove (51.31232,-56.708475)
ERROR: hidden waypoint colocated with visible, ca.ca014trknew CA14 (34.335955,-118.508025) and ca.ca014 +CA14Trk (34.335955,-118.508025)
ERROR: hidden waypoint colocated with visible, fl.fl200 +X008(US301) (29.7613,-82.141685) and fl.us301 CR1469/1471 (29.7613,-82.141685)
ERROR: hidden waypoint colocated with visible, az.us066hiskin +Coc/Yav (35.507566,-113.183813) and az.az066 Coc/Yav (35.507566,-113.183813)
ERROR: hidden waypoint colocated with visible, mo.us066hisstl FR144 (37.192801,-93.542061) and mo.mo266 +X01 (37.192801,-93.542061)

[si404]

There are currently 50 instances I detected of the situation I described in the OP.  Places where the same coordinates are used on two routes, either intersecting or as part of a concurrency, but one route's waypoint is hidden while the other is visible.  Thoughts?

There's a few I see as genuine errors, but most seem to make narrative sense.
Eg I-22 ending at I-269, but the latter isn't built yet so isn't shown on US78. Ditto border non-interchanges, freeway temporary termini, etc.
Or where exits were closed on the E15 when the A1 was converted to A1(M) - these closed exits have nothing to do with the A1(M) and so are hidden as clutter, but they do have relevance for the E15.
Or where there's two interweaving non-concurrent routes that you want to map properly and so a collector-distributor arrangement has the same points on all routes (with one set of carriageways shifted by 1 on every other point to stop the concurrency) in order to map properly, rather than having the carriageways with fewer junctions keep cutting about. Look at Bologna on CHMSys mapping and even on that scale you can see that 'TanBol' (now RA1) and A14Cas don't seem to take the same path. Now look at it on TM - the routes are hard to tell apart, which is right as they have the same centreline, even if they share no pavement (and you clinch the outer carriageways your trace will be cyan, whereas if you clinch the inner ones it will be blue).

[yakra]

OK, so I already have another example and this one might not be so simple.  US 1 in CT at its intersection with US 7 has 4(!) waypoints, which to me seems to violate the one point per interchange rule.  The immediate concern to me is the hidden US7 point in the middle of the mess, concurrent with US 7's US1 point.  It is not in use in TM, so I could just remove it, but perhaps we can get a few people to take a look at that area and decide if it makes more sense to have single US7 point, exactly where +US7 is now, and get rid of US7_S, US7_N, and RivAve (keeping US7_S as a hidden alternate for US7, as that point is in use).

This was one of the first routes I drafted, years ago. Every now & again I think of how I handled that interchange (It's what Jason/oxlahun, who drafted usact, would call "a bit rampy"), and it bugs me. Though I've never gotten around to actually fixing it. And yes, the plan was to collapse all 4 into a single point. So I'll put in a pull request to fix that; the heck with further looking around. I was a bit hesitant at first, thinking RivAve was a numbered route, but now I see that's not a problem.

-----

To get back to the question in the OP, a couple examples came to mind...

The US6 / RI10 concurrency is kept intact via a little gimmickry: WesSt is only accessible from RI10, and Bro is only accessible from US6. They have the same coords, however: RI10 uses standard waypoint positioning at the crossroad, and US6 has its point at the gore where the ramp departs the mainline. Voila, the concurrency is preserved.
Although... This could cause some of the same wacky hijinks as with US277 below (read on)...
"If ... a traveler uses it as a waypoint in one of the other routes, multiplex detection would kick in and show... An' it don't work like that!"
So perhaps it's best to just 1PPI away Bro/WesSt... Thoughts?

In Wichita Falls, TX, I considered having a 15thSt point visible on US281/287 and hidden on US277, and decided against it.
I typed out my reasoning on the old forum:

Check out the southern split of US 277/281/287.
Let me draw your attention to waypoint 14th/15thSt.

14th & 15th are both two-way streets.
So this 1/2-diamond should be recentered at 15thSt. Except...
Due to the way the ramps are set up here, there's access to/from US281/287, but not US277.

A couple things I could do:
* Hide the point on US277, in order to not break the multiplex.
* Decide that with this interchange entirely within the footprint of the `plex-split interchange, to just fold it into such, and delete this point. (Recentered @15thSt, the two points are 0.21 mi apart.)

Leaning slightly towards the latter, ATM. Thoughts?

Thinking about it some more...
If 15thSt is hidden in US277, and a traveler uses it as a waypoint in one of the other routes, multiplex detection would kick in and show their travel on US277 ending there. An' it don't work like that!
Seems the thing to do is go with the 2nd option, and delete the point.

So I guess the idea of having hidden+visible points colocated didn't quite sit right with me. And if multiplex detection can cause wacky hijinks with a user's map, as in that last example, all the better to avoid it.
I can't really think of any legitimate reason to have a point visible on one route & hidden on another. It seems that such cases could be easily 1PPI-ed away.

Eg I-22 ending at I-269, but the latter isn't built yet so isn't shown on US78.

Yes, this case. This is a legitimate reason, IMO.

ERROR: hidden waypoint colocated with visible, nl.nl430 +Goo/Yah (51.31232,-56.708475) and nl.nl430 MisCove (51.31232,-56.708475)
ERROR: hidden waypoint colocated with visible, me.me170 StaFacRd (45.421144,-68.137025) and me.me169 +x0StaFacRd (45.421144,-68.137025)

I'll fix these.

[yakra]

Should I remove Bro from RI US6, and WesSt from RI10?
Thoughts?

[Jim]

I am trying out some new datacheck errors in the test database, some of which are related to this thread.  Please have a look and let me know if these seem useful enough to be able to track them down.  Don't worry about marking false positives yet but by all means feel free to fix legitimate errors.

http://tm.teresco.org/devel/datacheck.php?dbname=TravelMappingTest

The relevant entries are currently marked as "VISIBLE_AND_HIDDEN_C" but I have renamed to "VISIBLE_HIDDEN_COLOC" for the next run.

[si404]

While you are on datacheck, can checks for visible points in the form X123456 be done (needs more than three digits due to Chinese X123 roads) as when making things like Asian Highways (almost always) and there's a manditory point (GSJ usually) that I can't work a label out for (or I thought I'd come back to name it but overlooked) I've just removed the '+' from the beginning of the label.

[Jim]

While you are on datacheck, can checks for visible points in the form X123456 be done (needs more than three digits due to Chinese X123 roads) as when making things like Asian Highways (almost always) and there's a manditory point (GSJ usually) that I can't work a label out for (or I thought I'd come back to name it but overlooked) I've just removed the '+' from the beginning of the label.

So a check for any visible label with X followed by some number of digits?

[yakra]

Six digits, I would say. It's the standard format for hidden points in WPTedit.

I second Si's proposal; it would be useful for me too.
When developing Canadian systems, sometimes I'll remove the + from the beginning of a shaping point, so that the coords will snap to the nearest road junction (a feature that ignores shaping points) when I run the route file thru GISplunge. (That way, the point's right where I need it if I ever discover a name for it later.)
Sometimes, I can potentially forget to re-add the + to the label before committing the file.

[Jim]

OK, the new datacheck as requested matches labels that are X followed by 6 digits.  I'll try a run that will go to the TravelMappingTest database first.

[Jim]

OK, the new datacheck as requested matches labels that are X followed by 6 digits.  I'll try a run that will go to the TravelMappingTest database first.

And it's live in the TravelMappingTest DB. 

http://tm.teresco.org/devel/datacheck.php?dbname=TravelMappingTest

Please ignore the fact that these new datachecks are in red.  That's just the default.

[si404]

Thanks, that's great!