Travel Mapping
Highway Data Discussion => In-progress Highway Systems & Work => Completed Highway Systems Threads => Topic started by: yakra on January 19, 2016, 12:21:47 pm
-
Old forum thread:
http://clinched.s2.bizhat.com/viewtopic.php?t=1688&mforum=clinched
I'll continue work on the next block of KY state highways, 1000-1499.
Same system and same sources as in the lower-numbered routes
http://clinched.s2.bizhat.com/viewtopic.php?t=1530&mforum=clinched
-
List of unsigned highways:
KY3S
KY8C
KY8S
KY15C
KY15S
KY66C
KY90S
KY94C x2
KY189C
KY194S
KY211 Colfax section
KY414
KY439C
KY451C
KY471
KY748
KY788
KY887
KY929
KY938
KY971
KY1001
KY1074
KY1082
KY1130S
KY1225
KY1174
KY1341
KY1354
KY1381
KY1452
-
This subsystem of highways is ready for peer review.
(emphasis added)
-
I'll have a look later.
-
Looked at the 1000-1025, no problems there.
This subsystem must have been tedium! I got bored very quickly, hence why I've only looked at 25 routes (1/20 of the way there if I look at them all).
And when I see routes in the 3000s, goodness me...
-
Thanks, Si. I figured the British A Roads would be rather similar.
I realized that I forgot to make the connected routes file, so I will do that shortly.
I have a rule I have followed for years that rarely needs to be invoked, which is to allow at least 2 weeks for collaborators to look over a system in the peer review phase, even if one review is done in less than 2 weeks. So anyone can still chime in if they would like before this subsystem is activated.
-
Looked at 1075-1099, 1150-1173, 1226-1249, 1300-1324, 1375-1399, 1450-1473 (every 3rd block of 25)
KY1156 - US25 -> US25/421
KY1165 - KY7_S <-> KY7_N - the directional labels are the wrong way around
KY1247 - US27_* -> town names better than letters from A to N? Might not be possible
KY1303 - KY1829_W on Industrial Road according to tile mapping.
KY1312 - Cum/Met not on county boundary from open source data (MQ Open/Mapnik)
KY1313 - Rus/Ada not on county boundary from open source data (MQ Open/Mapnik)
KY1316 - MosCreRd -> KY2352
KY1377 - KY80 -> KY80/1371 (and a similar move on KY1371)
KY1384 - US23/119_N - doesn't meet US119, does meet US460
KY1460 - points need moving onto the highway (and on interchanging roads); MayoTr_N -> KY3495; hidden point moved to nearby road?
General thing - is there too much shaping with non-mandatory and hidden points?
-
Are these datacheck errors false positives?
KY KY1013 KY3303 KY32_W KY32_E sharp angle 144.3 deg.
KY KY1146 KY476/550_N KY80_E KY80_W sharp angle 138.66 deg.
KY KY1165 MacBraRd_N LeftForkRd MacBraRd_S sharp angle 144.65 deg.
KY KY1262 KY1900 KY1707 +X800335 sharp angle 140.78 deg.
KY KY1329 +X480159 +X341250 LeeFauRd sharp angle 139.86 deg.
KY KY1336 End +X317930 KY697 sharp angle 147.43 deg.
KY KY1426 KY2030 KY680_E KY680/979 sharp angle 156.87 deg.
KY KY1461 HicLn +X855386 KY1461_N sharp angle 163.16 deg.
KY KY1461 KY1461_N KY1461_W 36.524501,-88.914328 duplicated coordinates.
KY KY1461 KY1461_N label references own route
KY KY1461 KY1461_W label references own route
KY KY1499 KY366 +X633272 +X817752 sharp angle 135.27 deg.
-
KY 1051: Should KY448 point be aligned on the KY 448 mainline (and the corresponding point in that route?)
KY 1054: reposition KY330_S?
KY 1059: Google and Yahoo show the route continuing to KY 32, but expect you looked into this when plotting
KY 1070: Would the road name be an acceptable alternative to "End"?
KY 1072: Point at Montague Rd for travelers who might be on this only to access the park?
-
bump - I know you have a lot on, but the changes haven't been made. Consider this a gentle reminder that this subsystem is waiting here.
-
New discussion:
Based on Tim's statement that this system is ready for peer review, I'll be promoting it to preview status shortly.
I may still handle the lingering issue of KY1076 at the TN border (http://tm.teresco.org/forum/index.php?topic=19.0). (Edit: I now have a pull request in for this.)
mapcat, do you have any interest in checking out the above suggestions, making any necessary edits/corrections/changes, and seeing this subsystem thru to activation?
-
Sure, I'll get on it.
-
I've made it through the 10XXs and found a few others that are unsigned:
1001*
1018
1052
1074*
1082*
1085
Those marked * are shown on KYTC county maps but GMSV does not show any signs. KY1001 has a 0 mile marker at its intersection with KY80, but that's the only evidence that it exists (no GMSV coverage on the route itself).
A few others have minor problems (waypoints slightly off center or in need of renaming):
1004 (rename waypoint, move CroCreRd)
1043 (rename waypoint(s))
1044 (rename waypoint(s))
1049 (rename waypoint(s))
1050 (rename waypoint(s))
1053 (rename waypoint(s))
1054 (move KY330_S)
1055 (rename waypoint(s))
1064 (rename waypoint(s))
1071 (rename waypoint(s))
1075 (move ends slightly)
1083 (rename waypoint(s))
1087 (rename waypoint(s))
1088 (rename waypoint(s))
1089 (rename waypoint(s))
1096 (move BigCrkRd slightly)
1098 (rename waypoint(s))
Tim evidently was using OSM data for waypoint names, but it seems to me that KYTC names would be more appropriate. Is there any precedent?
Others need additional waypoints:
1027 (KY756)
1057 (KY2001)
Finally these need more significant work:
1020 has very spotty signage throughout urban Louisville, but KYTC's route is close to what we have. Also, KYTC shows part of it along a one-way pair, so waypoints need to be re-centered.
1081 definitely ends at KY40 per both GMSV and KYTC.
KY 1051: Should KY448 point be aligned on the KY 448 mainline (and the corresponding point in that route?)
KY 1054: reposition KY330_S?
KY 1059: Google and Yahoo show the route continuing to KY 32, but expect you looked into this when plotting
KY 1070: Would the road name be an acceptable alternative to "End"?
KY 1072: Point at Montague Rd for travelers who might be on this only to access the park?
I think 1051 is probably ok as-is (the "mainline" intersection is more of a ramp now), but agree on 1054, and could be swayed on 1059 (it seems to end at KY32 according to GMSV but KYTC agrees with what we currently have). It looks like KY1070 isn't also named Howell Rd, so I can change End to HowRd; looks like I ought to do the same with KY1094. I can add a waypoint for 1072 at the park entrance.
I'll make the edits and submit the pull request sometime this weekend.
-
Tim evidently was using OSM data for waypoint names, but it seems to me that KYTC names would be more appropriate. Is there any precedent?
That shouldn't be a problem. They have been credited on the old 'Sources and Links' page for the project. Especially since I was the one the created the original KY US Highway files (& Parkways).
http://cmap.m-plex.com/docs/sources.php
-
Tim evidently was using OSM data for waypoint names, but it seems to me that KYTC names would be more appropriate. Is there any precedent?
I might do some GMSVing around until I get a good feel for what's more likely to match what's in the field.
In my very limited experience at the south end of KY US127BusAlb, KYTC names were more accurate.
-
I might do some GMSVing around until I get a good feel for what's more likely to match what's in the field.
In my very limited experience at the south end of KY US127BusAlb, KYTC names were more accurate.
Whenever there's a discrepancy I check GMSV. Trouble is that much of the GMSV in the eastern part of the state is very old and street signs are illegible.
-
1020 has very spotty signage throughout urban Louisville, but KYTC's route is close to what we have. Also, KYTC shows part of it along a one-way pair, so waypoints need to be re-centered.
I'd go with KYTC... Is there any signage you can find that outright contradicts this?Or is signage just... poor?
KY1059: extend to KY32 or not?
I'll let you be the judge of that. Anyone else got an opinion?
If this were Maine, I'd include the extension. Probably. But this isn't Maine, is it?
-
1020 has very spotty signage throughout urban Louisville, but KYTC's route is close to what we have. Also, KYTC shows part of it along a one-way pair, so waypoints need to be re-centered.
I'd go with KYTC... Is there any signage you can find that outright contradicts this?Or is signage just... poor?
Louisville map (http://transportation.ky.gov/Planning/SPRS%20Maps/Louisville_city.pdf) (see sheet 3)
Signage exists at the intersection with US60A but is (almost) entirely missing north of it. Between there and I-264 there isn't any signage but it follows one roadway without turning. South of I-264, it turns several times through a residential area but the first signed turn I found is at the intersection of Southside Dr and National Turnpike, even though it has a rather convoluted routing along residential streets in between. From that point southward, it seems to be signed well.
I examined some more logical routes for it to take through the city but didn't find signs anywhere else nearby, either.
Since it's signed at US60A, I think we should at least keep it from that point south. North of there, its official route at least follows the same streets (2nd & 3rd) to its unmarked end, and according to some of my old official maps it has followed them since at least the early '90s (older maps didn't have as much detail). OTOH, the US routes and KY61 and KY864 are comparatively well-signed downtown. Right now I'm inclined to reroute it along the KYTC routing all the way to its official end, since I can't find any signage that conflicts with that, although deleting it north of US150 (the first signed route it encounters north of US60A) would make about as much sense.
-
Here are the notes for the 11XX set:
Deleted or unknown:
1111
1143
1188
1190
1197
Unsigned but still KYTC-maintained:
1174
Incorrect endpoint locations:
1106
1145
1185
Incorrect endpoint/waypoint names:
1101
1102
1103
1110
1116
1117
1120
1122
1124
1132
1133
1140
1147
1148
1150
1151
1152
1153
1156
1165
1167
1169
1171
1175
1179
1181
1186
1187
1192
1198
Missing or extraneous waypoints (false intersections that can be eliminated without affecting shape):
1107 (missing KY1145)
1151
1153
The rest appear ok. Corrections will be made and files submitted over the weekend.
I have not begun checking for datacheck errors because I am unsure how to mark false positives.
-
I have not begun checking for datacheck errors because I am unsure how to mark false positives.
The code in the "FP Entry to Submit" column gets added as a new line to datacheckfps.csv (https://github.com/TravelMapping/HighwayData/blob/master/datacheckfps.csv) in the root directory of the HighwayData repo. (The URL will be different when pointing to whatever branch you're editing.)
-
I have not begun checking for datacheck errors because I am unsure how to mark false positives.
The code in the "FP Entry to Submit" column gets added as a new line to datacheckfps.csv (https://github.com/TravelMapping/HighwayData/blob/master/datacheckfps.csv) in the root directory of the HighwayData repo. (The URL will be different when pointing to whatever branch you're editing.)
A note on this: as long as we keep it more or less alphabetized different people's edits there will not conflict with each other, so it's safe to submit pull requests with commits to datacheckfps.csv. Of course you can also put them in the text description of the pull or email them and I can add them. Just pointing out that this situation is different from updates.csv, where we'd all be adding in the same place, which an automatic merge would not be able to handle (nor should it be able to) so we'd end up with conflicts to resolve when merging in those pull requests.
-
The code in the "FP Entry to Submit" column gets added as a new line to datacheckfps.csv (https://github.com/TravelMapping/HighwayData/blob/master/datacheckfps.csv) in the root directory of the HighwayData repo. (The URL will be different when pointing to whatever branch you're editing.)
Um...
When I look at the file the columns are:
Route;Waypoint1;Waypoint2;Waypoint3;Error;Info
-
Notes for the 12XX routes, to be fixed over the weekend:
Deleted or unknown:
1264
Unsigned but still KYTC-maintained:
1225 (signed as "To 80"
Incorrect endpoint locations:
1201
1202
1205
1206
1227
1240
1242
1246
1258
1271
1273
1275
1281
1283 (end marked at 45; KYTC says it ends 1 block east)
Incorrect endpoint/waypoint names:
1203
1214
1220
1223
1230
1232
1234
1238
1241
1244
1245
1247 (town names for KY1247 aren't going to work since there are more intersections with US27 than there are towns)
1249
1250
1260
1267
1272
1273
1274
1277
1292
1295
1298
Missing or extraneous waypoints (false intersections that can be eliminated without affecting shape):
1268
Incorrect shaping points
1221
-
Notes for the 13XXs, to be fixed this weekend:
Deleted or unknown:
1334
1341
1358
1370
1378
1386
Unsigned by KYTC-maintained
1381
Incorrect endpoint locations (most are small adjustments)
1308
1313
1318
1335
1345
1346
1347
1361
1371
1384
1388
1391 (both ends)
1392
Incorrect endpoint/waypoint names:
1303
1305
1311
1316
1326
1327
1328
1329
1331
1332 (also needs waypoint for US231)
1336
1345
1348
1353
1355
1356
1359
1363
1366
1367
1373
1379
1380
1384
1385
1389
1393
1396
1399
Reverse order to match mileposts
1301
1355
1399
Missing or extraneous waypoints (false intersections that can be eliminated without affecting shape):
1336
1340
Other
1309 (move shaping point to nearby BMarRd)
1376 (move FarRd)
KY1303 - KY1829_W on Industrial Road according to tile mapping.
KY1312 - Cum/Met not on county boundary from open source data (MQ Open/Mapnik)
KY1313 - Rus/Ada not on county boundary from open source data (MQ Open/Mapnik)
KY1316 - MosCreRd -> KY2352
KY1377 - KY80 -> KY80/1371 (and a similar move on KY1371)
KY1384 - US23/119_N - doesn't meet US119, does meet US460
1303: noted above
1312 & 1313: the endpoint matches the endpoint (county line) indicated in KYTC aerials, so this is unnecessary
1316: KY2352 is unsigned, so this is unnecessary
1371 & 1377: these routes don't intersect KY80 in the same location, as indicated in the KY80 file, so this is unnecessary
1384: noted above; the KY1460 end is also at the wrong location
-
Notes for the 14XXs, which I hope to fix this weekend.
Deleted or unknown:
1423
1467
1474
1478
1480
1493
Unsigned but KYTC-maintained:
1452
Incorrect endpoint locations (most are small adjustments):
1436
1439
1450
1456
1457
1461
1471
1475
1476
1479
1483
1499
Incorrect waypoint locations/missing waypoints (including shaping points to convert to waypoints):
1412
1414
1426
1428
1428Pai
1429
1453
1459
1460
1463
1469
1471
1473
1482
1491
1494
1496
1499
Incorrect endpoint/waypoint names:
1401
1411
1414
1415
1418
1426
1428
1439
1441
1442
1446
1448
1450
1453
1458
1460
1464
1465
1470
1472
1482
1486
1489
1492
1494
1496
1498
Reverse order to match mileposts:
1421
1491
All others appear to be in good shape.
KY1460 - points need moving onto the highway (and on interchanging roads); MayoTr_N -> KY3495; hidden point moved to nearby road?
These changes are included above.
-
KY1486 has a connector to KY17 (officially called KY1486C) marked as KY1486, which leads to a strange 3-way intersection where every direction is KY1486 (https://www.google.com/maps/place/State+Hwy+1486,+Independence,+KY+41051/@38.9709258,-84.5296763,3a,75y,59.44h,90t/data=!3m7!1e1!3m5!1s0lPKjptGL3cTRy-DBdKlpQ!2e0!6s%2F%2Fgeo3.ggpht.com%2Fcbk%3Fpanoid%3D0lPKjptGL3cTRy-DBdKlpQ%26output%3Dthumbnail%26cb_client%3Dmaps_sv.tactile.gps%26thumb%3D2%26w%3D203%26h%3D100%26yaw%3D52.78828%26pitch%3D0!7i13312!8i6656!4m2!3m1!1s0x8841b959823d81b5:0xacc17177309d43ec). I'm inclined to leave it out, but am interested in hearing reasons for including it as well. There's at least one connector in the state highway system marked with a "C" sign (https://www.google.com/maps/@37.9831236,-85.7173302,3a,56.4y,36.18h,85.5t/data=!3m6!1e1!3m4!1sLvpn4jmHniGmkYeia-ZvRg!2e0!7i13312!8i6656), which tells me that if it were important enough, 1486C would have one as well (assuming they could fit one onto the sign).
-
KY1486 has a connector to KY17 (officially called KY1486C) marked as KY1486, which leads to a strange 3-way intersection where every direction is KY1486 (https://www.google.com/maps/place/State+Hwy+1486,+Independence,+KY+41051/@38.9709258,-84.5296763,3a,75y,59.44h,90t/data=!3m7!1e1!3m5!1s0lPKjptGL3cTRy-DBdKlpQ!2e0!6s%2F%2Fgeo3.ggpht.com%2Fcbk%3Fpanoid%3D0lPKjptGL3cTRy-DBdKlpQ%26output%3Dthumbnail%26cb_client%3Dmaps_sv.tactile.gps%26thumb%3D2%26w%3D203%26h%3D100%26yaw%3D52.78828%26pitch%3D0!7i13312!8i6656!4m2!3m1!1s0x8841b959823d81b5:0xacc17177309d43ec). I'm inclined to leave it out, but am interested in hearing reasons for including it as well. There's at least one connector in the state highway system marked with a "C" sign (https://www.google.com/maps/@37.9831236,-85.7173302,3a,56.4y,36.18h,85.5t/data=!3m6!1e1!3m4!1sLvpn4jmHniGmkYeia-ZvRg!2e0!7i13312!8i6656), which tells me that if it were important enough, 1486C would have one as well (assuming they could fit one onto the sign).
I see it colored green on the Kenton county SPRS map. (No caption though; too small/short.) So that means it's part of the state system, I think. (Right?) That combined with the signage would make me lean toward including it if I were drafting USAKY3. Signed as such, I'd probably choose KY1486 + a City suffix over calling it KY1486C, whatever it may be designated internally. Though, what to pick for said suffix? I'm glad I have the luxury of not having to ponder this too deeply. :D I don't know, man, I didn't do it. My feelings aren't terribly strong in any particular direction.
-
@yakra: if you don't think it's too short, then I'll have a look-see how many of these there are out there of at least approximately the same length and bring in the whole lot of them at once.
Speaking of bringing in the whole lot: since usaky3 was complete when I inherited it and I conducted what I believe to be a thorough evaluation of all nearly 500 routes, making necessary corrections, can we promote this set to Active? :)
-
Any further review needed before activating the set?
-
Speaking of bringing in the whole lot: since usaky3 was complete when I inherited it and I conducted what I believe to be a thorough evaluation of all nearly 500 routes, making necessary corrections, can we promote this set to Active? :)
Personally, I'd be okay with that. Well... on further thought, I'm waffling back & forth a bit...
On the other hand, as a new collaborator, a little checking wouldn't hurt, to make sure conventions are followed, point labels and positioning are A-OK, and the like.
On the third hand, for this reason, I kept a pretty close eye on your commits, and saw very few problems. I pointed out a few things, and you either fixed them or convinced me that your approach was correct. So yeah -- we're probably OK here, though another pair of eyes wouldn't hurt.
However, if you're going to go ahead with Operation Bring In the Whole Lot as described in your first paragraph, maybe it's better to hold off till that's done.
Anyone else have any Ayes or Nays, comments?
-
Having given review to this system back in the CHM days, I don't want to put effort into reviewing it now - I gather that my comments from then were dealt with, so I'm fine with activation whenever yakra feels its OK.
This one remaining datacheck error looks to be a false positive.
ky.ky1499;KY366;+X633272;+X817752;SHARP_ANGLE;138.61
-
This one remaining datacheck error looks to be a false positive.
ky.ky1499;KY366;+X633272;+X817752;SHARP_ANGLE;138.61
Thanks for pointing out that this one is back from the dead. I had added this error to the fps file already, but it came back after I adjusted a waypoint location and the angle changed slightly (the last 3 digits were different). Is there a reason that the error log has to state the angle measurement?
-
2 weeks with no further comments...can we activate?
I'll deal with the connector routes separately, since I plan on going through the KYTC material county by county collecting the remaining routes before making new files, and it will be easier to look for other routes ending in "C" while doing that.
-
I personally don't have a strong opinion either way on including the connector routes, and won't shout you down if you want to activate usaky3.
Do you have any knowledge as to whether similar routes were included/excluded from Kentucky Highways 1-999?
You're probably aware, but I just wanna make sure, that on activation, all the usaky3 files would be moved into the usaky directory, and the contents of the usaky3 CSVs would be merged into the usaky CSVs.
-
Do you have any knowledge as to whether similar routes were included/excluded from Kentucky Highways 1-999?
The only ones were the routes actually signed KYXXXC.
You're probably aware, but I just wanna make sure, that on activation, all the usaky3 files would be moved into the usaky directory, and the contents of the usaky3 CSVs would be merged into the usaky CSVs.
I realized that would happen, and was assuming that Jim would have a simple way of doing it behind the scenes. Or should I make the moves myself?
-
You're probably aware, but I just wanna make sure, that on activation, all the usaky3 files would be moved into the usaky directory, and the contents of the usaky3 CSVs would be merged into the usaky CSVs.
I realized that would happen, and was assuming that Jim would have a simple way of doing it behind the scenes. Or should I make the moves myself?
Right - if you're not comfortable with git from the command line and similar things, it's probably easiest to let me know when it's ready and I'll do all of the appropriate renaming and merging.
-
I realized that would happen, and was assuming that Jim would have a simple way of doing it behind the scenes. Or should I make the moves myself?
It's really simple anyway - copy the data in the .csv files of usaky3 (minus the header line) and paste into the respective usaky .csvs at the bottom. Do a find-replace 'usaky3' for 'usaky' and then that's it.
Though, of course, you have to also move the files from KY/usaky3 to KY/usaky, which might prove harder for you. You don't need git from the command line to do it. The 'for Windows' makes it rather easy (simply do it the same way you'd move any bunch of files) and I believe the web interface now allows you to drag and drop to add files to folders (though I don't know whether you can do mass deletions).
-
All right, I'll try it and let you know if I run into any problems.
-
No problems with changing .csv files.
But now that I'm moving the individual .wpts from usaky3 to usaky, I've run into this error message in my local fork:
Sorry, we had to truncate this directory to 1,000 files. 166 entries were omitted from the list.
Does git have a limit of 1000 files in a directory? Why else might I be getting this message?
-
Does git have a limit of 1000 files in a directory? Why else might I be getting this message?
My guess is that the web interface is just showing 1000 but that they all exist.
-
My guess is that the web interface is just showing 1000 but that they all exist.
Exactly this.
-
OK, thanks. Submitted.
I did not attempt to delete the files from the usaky3 directory yet since I want to make sure all of them copied correctly first.
-
The usaky3->usaky merge is complete and usaky3 files are now deleted.
Any objections to my system updates entry for this? I marked usaky3 as "merged" and usaky as "extended".
-
The usaky3->usaky merge is complete and usaky3 files are now deleted.
Any objections to my system updates entry for this? I marked usaky3 as "merged" and usaky as "extended".
Thanks! The updates language makes sense to me.
-
I marked usaky3 as "merged" and usaky as "extended".
Great idea! :)
btw: Si merged some systems last year (Spain, Italy, UK and Czech Republic). The "New Status" is cut: "delete (merge in ? ? ? ?"...
-
btw: Si merged some systems last year (Spain, Italy, UK and Czech Republic). The "New Status" is cut: "delete (merge in ? ? ? ?"...
merged seems better
-
btw: Si merged some systems last year (Spain, Italy, UK and Czech Republic). The "New Status" is cut: "delete (merge in ? ? ? ?"...
merged seems better
Please feel free to update old update entries as you see fit.