canmb: Manitoba Provincial Trunk Highways

[yakra]

Tue Nov 03, 2015 1:30 pm on CHM forum
Canonical Shapefiles:
NRN_MB_6_0_ROADSEG, dated 2016-06-06
(As opposed to the 6.0 shapefiles dated 2013-04-11 or 2013-11-04...  )

This post uses used nonstandard color coding:
red:   rough draft done & under review
mgta:   Now: ready for gisplunge / Was: ready, except nomenclature not sorted
prpl:   Now: gisplunged; needs review / Was: nomenclature sorted; ready for GISplunge
blue:   ready to commit
cyan:   committed
grn:   live in hb

Route List
CANSPH:
Batch 1: 6 10 29 59 60 75 83 101
Children:
Batch 1: 10ADau 10AEth 10ASwa 10AFli 83ASwa
The Rest:
Batch 2: 2 3 3ACle 4 5 5ADau 7 8 9 9ASel
Batch 3: 11 12 13 14 15 17 18 19
Batch 4: 20 20ADau 21 22 23 24 25 26 27
Batch 5: 30 31 32 34 39 41 42 44 45 49
Batch 5: 50 52 57 67 68 77 89 110 190

Each batch: save all files, to ensure terminal CRLF

ToDo / Notepad / Checklist

##A routes: NH-style or ME-style WRT cities and suffixes? (ME style.)
Final Cleanup Checklist

MB2:
Check for road names:
+X473519 -> 164W
+X122992 -> 38N
+X991418 -> 152W
+X380889 -> 150W
+X234613 -> 42N
+X866568 ->39N
+X507316: move NE to GilSchRd, Culross
+X742347 -> 50N (replaced; SegDumped)
+X212134 -> 12W
+X982131 -> 9W
MB340: This *is* 340 and not 344, right? (Right!)

MB3:
gisplunge SK/MB. Or QGIS it.
Check for road names:
+X1 -> 159W
+X426434 -> 18N
+X942566 -> 15N
+X475612 -> 56W
+X350035 -> 46W_N
+X338928 -> 46W_S
MB458: Is the west one 458 at all? (See also MB21 MB23) GIS and GMSV say no. Renamed 153W.
MB458: Is the east one 458 or 340? (458!)

MB5:
Check for road names:
BisJump -> 84W
+X669167 -> 83W
34N: Not MB235 per GMSV, GIS, hwy map
+X107847 -> 36N
+X750452 -> 42N
75N: Not MB465 per GMSV, OSM, GIS, hwy map
+X799794 doesn't exist!
+X764749 -> 88W
+X641020 ->106W
+X408505 -> 141N
+X341469 -> 146N
+X218666 -> 139W
SarCSRd -> DanSpotRd
Delete MouAve on 16-plex? (Hell, it can stay.)
MB482: Right province? Single point with border? (Right province. GIS, Bing, and OSM agree with my coords. Two points. Signage from the MB side shows SK10 continuing west, and MB5 only to the right from NB MB482. Prolly a matter of convenience & simplicity. The weird bit is that per GIS, MB ends at MB482. The road to the border has no RTNUMBER1, and L_STNAME_C = "10 Highway". I'm pretty confident this is an error.

MB6:
GIS: Does Railway Ave connect, near 1St? Yes; replaced.
Possible realignments, SegDump oldest shapefiles:
• WilLakeRd +X71 (out of tolerance) / 63eeba5f7e9a4db19c6d3b6dd7be772e-12177.wpt = Paydirt
• +x82 / 63eeba5f7e9a4db19c6d3b6dd7be772e-12176.wpt = Paydirt
    ~ +X385446 will not gisplunge; use OSM
• +x108 (or is it rail? check old maps; MyTopo) / Nothing.
• S of SealRd (or is it rail? check old maps; MyTopo) / Nothing.
NMPs: MB67 MB411

MB8: Check north end in GIS. (Extended!)

MB10:
MB5_E & MB5_W -> MB5_S & MB5_N? (Check GMSV.) (E&W.)
MB10A/83 label OK? (OK.)
realignments / OldHwy points near +x46?
• Oldest shapefiles show modern alignment. No L_STNAME_C in latest files.
• MyTopo - inconclusive
RicAve via QGIS
+X7 -> 82N
MB355: coords from OSM
+x18 -> 102N
unhide +X30
NorShoRd: check GIS for name: None
LakeAudy: check GIS for name: None
Ochre River Trail, S of +X39
MoonLake
unhide +x1nww
194N
MinkCre -> 164N? check GIS for name
unhide +X880219
MB10ASwa: coords from OSM
unhide +X250842
BriRd
unhide +x74nww
unhide +x75nww (SnapJct -> a bit north) Never mind.
BatRd
MB287: coords from OSM
potential visible point S of RocLakeRd; check GIS. No luck.
MB10AFli_E: coords from OSM
MB/SK: verify in QGIS
GMSV: MB5_E MB5_W MinkCre 194N

MB10ASwa:
+X269441 -> 215N
Is that MB83 or MB83A? (83A)
Goddammit (All signage at MB10_S / MB83 shows MB10A)

MB12:
check out MB304/500 for useless-plex: None; both rtes terminate here. Label stays as is.
check out MB203/404 for useless-plex: Yes, per GIS and GMSV. Label stays as is.
reposition MB15/101 junction: MB15 jct is fine. MB12 goes nowhere MB101. This comment doesn't appear to be intended for another route, with nothing else intersecting MB15/101. Total Nonsense?

MB17: +X732221 -> 4W
MB19: Points in correct order?

MB20:
Points in correct order?
UkrRd: Is it MB273? (Probably not.) No. Renamed 168N.
+X824572 -> MalRd
+X305150: unhide/rehide
+X349966 doesn't exist!

MB21:
RivAve -> GovRdE
+X772061 -> GovRdW
+X234613 -> 42N
ToMB259 = Rivers Rd?
Is SouEndRd MB477? No. Renamed 93N.
Something about (old?) MB458? (See also MB3.) Routes nowhere near each other. Looks like this should have been put under MB23, and the "MB21" reference under MB3 was a typo.

MB23:
MB458: or is it 340? (See also MB3) 458 only.
+X421881 -> 28N
+X602663 -> 52W
+X553551 -> 51W
ArcRd: correct per GIS
ChMes: Not MB431, right? Right.
Fix MB59 jct in both files

MB25: +X436196 doesn't exist!
MB26: Mac -> 71N

MB29:
USA/MB -> ND/MB
Designation changes?
• GIS as-is. (May not be reliable)
• Hwy Map: southern bits of "MB75" are not PTH, but rather PR
• GMSV: southern bits of "MB75" confirmed signed as MB200
Wikipedia:
• "In 2012, PTH 75 was re-routed to connect directly with Interstate 29, thus eliminating PTH 29." (MB29)
• "The old stretch of PTH 75 through Emerson is now part of an extended Provincial Road 200." (MB75)
• (MB200)

MB32: +X978219 -> 21W

MB34:
+X942566 -> 15N
+X795598 -> 64W

MB39:
+X279846 isn't in the shapefiles
X809743: unhide/rehide
+X906577 isn't in the shapefiles

MB42:
+X846515 -> 96N
TheDr

MB44:
points to label near Rennie?
+X626995 WildGooPl CadLake: No luck; nothing to change
+X410507 doesn't exist!

MB50: +X361242 isn't in the shapefiles

MB59:
USA/MB -> MN/MB
+X197763 doesn't exist!
MB23
34N: L_STNAME_C = "303 Road"; RTNUMBER1 = "None"; not labeled on official hwy map
Rte17 for extra credit? No, not shown in GIS.
MB12_N: what does GIS look like here anyway? SegDump it!

MB68:
Alignment NE of MB278 is right, right? Right, per GMSV. OSM & GIS outdated.
+X230108 -> LonLake
+X910188 isn't in the shapefiles; not legible in GMSV
+X664429: no nearby names found; leave hidden
+X409429 -> DogCre
Old514 is signed as such in the field

MB75:
Removed from MB201; relocated onto southern extension of MB29
CarSt_S & CarSt_N: MB246? No. GIS has MB246 ending at Caron St.

MB83:
at MB10 Swan River: alignment, and... MB83A?
43Rd = ROAD 34 N? No, looks more like 44N.
66N: L_STNAME_C = "441 Highway"; RTNUMBER1 = "None"; not on official hwy map
OldMB589 -> 142N (L_STNAME_C = "589 Highway"; RTNUMBER1 = "None"; not on official hwy map)
OldMB591 -> 152N (L_STNAME_C = "591 Highway"; RTNUMBER1 = "None"; not on official hwy map)
Ken -> MainSt_Ken; MainSt -> MainSt_Bir
+x53 +x54
MB1 plex
MB16/TCHYel: synced to newly GISplunged TCHYel
realignment at MB16_W
+x164Rd

MB83ASwa:
Check GMSV for existence (clearly exists)
Check GIS for existence (bypass still not shown)
GISplunge: try RTNUMBER1 = 83

MB101:
MB221 -> 50
MB7 -> 60

MB190:
NE end? Exactly where I had it.
Is any of this in GIS yet? "Nothing yet, sir!"

[oscar]

MB83: at MB10 Swan River: alignment, and... MB83A?

MB83A (Swan River):
Check GMSV for existence (clearly exists)
Check GIS for existence

FWIW, not only did I see MB 83A signs in the field, on one of my two visits to Swan River in 2013 and 2015, I clinched the route.

[yakra]

Nomenclature
42W / Rd42W / Rd42 / 42Rd nomenclature
• does not begin with MB
• does not begin with +x or +X
• contains numeral(s)

Total sans TCH: 26 / 0 / 79 / 7
Total with TCH: 26 / 0 / 93 / 9

Look over false positives. Including 42Rd style labels, as these may not be survey line roads
Put the genuinely questionable ones in the notes in the OP

-----

TCHMai: 0 / 0 / 11 / 2
Rd161 137Rd Rd100 Rd78 Rd70 Rd64 Rd59 Rd28 Rd20 Rd19 32Rd Mile54 Rd26 Rd86

TCHPor: 0 / 0 / 1 / 0
Rd34

TCHYel: 0 / 0 / 2 / 0
Rd149 Rd50

MB2: 8 / 0 / 2 / 0
164W 38N 152W 150W 42N 39N Rd88 Rd50 12W 9W

MB3: 8 / 0 / 1 / 0
159W 153W 18N 15N 56W 46W_N 46W_S Rd34 19N

MB5: 10 / 0 / 6 / 0
84W 83W Rd17 Rd34 36N 42N Rd70 Rd75 Rd111 88W 106W 141N 146N 139W 146W Rd156

MB6: 0 / 0 / 2 / 1
Rd15 513Rd Rd1

MB7: 0 / 0 / 2 / 0
Rd101 Rd122

MB8: 0 / 0 / 2 / 0
Rd88 Rd136

MB9: 0 / 0 / 1 / 0
Rd107

MB10: 0 / 0 / 7 / 0
Rd137 Rd201 Rd204 Rd233 Rd247 Rd252 Rd253

MB12: 0 / 0 / 5 / 1
Rd77 Rd12 Rd54 402Rd Rd49 Rd25

MB13: 0 / 0 / 1 / 0
Rd54

MB14: 0 / 0 / 1 / 1
1Rd Rd11

MB15: 0 / 0 / 1 / 0
Rd59

MB17: 0 / 0 / 3 / 0
Rd94 Rd98 Rd109

MB19: 0 / 0 / 1 / 0
Rd90

MB20: 0 / 0 / 8 / 0
Rd142 Rd151 Rd184 Rd194 Rd195 Rd206 Rd127 Rd129

MB23: 0 / 0 / 4 / 0
Rd26 Rd25 Rd71 Rd41

MB26: 0 / 0 / 1 / 0
Rd30

MB31: 0 / 0 / 0 / 0
Mile8

MB34: 0 / 0 / 3 / 0
Rd19 Rd72 Rd77

MB45: 0 / 0 / 8 / 0
Rd120 Rd117 Rd112 Rd143 Rd111 Rd108 Rd130 Rd115

MB50: 0 / 0 / 4 / 0
Rd100 Rd105 Rd119 Rd84

MB59: 0 / 0 / 3 / 0
Rd34 Rd32 Rd34

MB68: 0 / 0 / 5 / 0
Rd23 Rd128_W Rd10 Rd128_E Rd129

MB75: 0 / 0 / 2 / 0
Rd2 Rd33

MB83: 0 / 0 / 6 / 4
Rd53 Rd54 33Rd 43Rd Rd66 Rd72 Rd73 164Rd 136Rd Rd202

[yakra]

MB83: at MB10 Swan River: alignment, and... MB83A?

MB83A (Swan River):
Check GMSV for existence (clearly exists)
Check GIS for existence

FWIW, not only did I see MB 83A signs in the field, on one of my two visits to Swan River in 2013 and 2015, I clinched the route.

Amusingly enough, MB83ASwa is the first file ready for upload.

[yakra]

TCH Cleanup

TCHBra:
recenter points, including MB10
only +X01 & 66thSt from GISplunge

TCHMai:
resync intersections with child routes
delete ThoDr?
nomenclature
MB100_W +-> 318 (it's signed now)
Check for other numbered exits
trimmed out overdense points in Winnipeg and Portage La Prairie
MB83_N recentered

TCHMin:
recenter a few points
GISplunge

TCHPor:
splunge
nomenclature
labels
endpoints

TCHWin:
Boom. Done.

TCHYel:
nomenclature
West of MB83
• just some shapers
MB83 plex
• MB83_N realignment (Edit OSM)
• MB45: RTNUMBER1 = 16
• The rest: RTNUMBER1 = 83
MB83 to MB10
• on
• MB21: coords from OSM
• on
MB10 plex
• MB10/16A: sync to TCHMin & MB10
• MB355: sync to MB10 (coords from OSM)
• 9thAve: RTNUMBER1 = 16
• MB10_S: sync to MB10
East of MB10
• Boom. Done.

[oscar]

Some quick comments on the initial in-dev canmb routes just added to the HB. I focused on the endpoints of the routes I've traveled (all but MB10AEth and MB60), to update my draft list file entries file. I didn't check intermediate points, other than some on MB 83 for which my coverage is incomplete.

MB10AFli -- MB10_W and MB10_E => MB10_N and MB10_S, to reflect the nominal direction of MB10 as it loops around Flin Flon

MB75 -- MN/MB => ND/MB

[yakra]

MB10AFli -- MB10_W and MB10_E => MB10_N and MB10_S, to reflect the nominal direction of MB10 as it loops around Flin Flon

Aah, but MB10A itself is east-west!

MB75 -- MN/MB => ND/MB

MB29:
USA/MB -> ND/MB

Now how did I manage that?... 
Fixed in my HighwayData repo.

[yakra]

GISplunge Bugs
This all worked perfectly well for years with my Ubuntu 10.04 compile, mumble grumble...

Last line not added to output file
In batch 2, affects mb.mb003acle.wpt, mb.mb004.wpt, and mb.mb009asel.wpt
These files are unmodified since Thanksgiving 2013.
This is likely a Unix/DOS terminating line feed issue.
1.) Check original drafts for terminating line feeds in hex editor. (Yup. They're missing.)
2.) Re-saving these files in Pluma makes things work, and is a suitable workaround. No DIFF from 10.04 compile.

Segfault in CSV mode
Does mb.mb009asel.wpt trip it up specifically? (No.) Or, is it an issue upon reaching EOF? The latter? (Yes.)
adding an extra line of whitespace @ EOF does not help

[oscar]

One comment on the second batch of files, re: north end of MB 8:

When I went up there in 2015, I saw an End MB 8 marker a few hundred feet north of the northern access to Hecla village, well short of road's end where you now have the route ending. A blurry GMSV image confirms my sighting. The End sign is around where the speed limit, and road quality, drop below rural PTH norms. So truncate MB 8 to Hec_N?

BTW, I saw no name on either Hecla access road, so I can't improve on Hec_N and Hec_S.

[yakra]

One comment on the second batch of files, re: north end of MB 8:

When I went up there in 2015, I saw an End MB 8 marker a few hundred feet north of the northern access to Hecla village, well short of road's end where you now have the route ending.

Interesting. My original draft of MB8 from Thanksgiving 2013 ended at Hec_N (maybe cuz OSM has the transition from the thicker orange secondary road trace to the thinner white local road trace only a tiny bit north of there?), but I wasn't 100% confident in that (maybe cuz Google showed the route continuing on farther?), and made a note to "Check north end in GIS". The shapefiles showed it going on to the end of the road in Gull Harbour, so I extended it.

A blurry GMSV image confirms my sighting.

Eh, it's not that blurry!

So truncate MB 8 to Hec_N?

Not really being able to pin down a more precise endpoint, and the original file having ended here, your suggestion sounds good.
Truncated in my HighwayData repo.

BTW, I saw no name on either Hecla access road, so I can't improve on Hec_N and Hec_S.

Nothing in GMSV; nothing in GIS.

[yakra]

Final Cleanup Checklist

Upload: 39
TCHMai 2 3 4 5 6 10 14 23 77 83 101

SK/MB coords: all but MB57
2 3 5 10 49 77

Labels containing '/' -- PTH/PR intersections: MB69/666
mb.mb002   MB22/250 -> MB22
mb.mb003   MB31/240 -> MB31
mb.mb014   MB30/332 -> MB30

Streets & Avenues: cardinal or ordinal? (MB10 has one of each)
mb.mb010   17St -> 17thSt
mb.mb083   2Ave -> 2ndAve

Look at OldMB##_X suffixes if I really want extra credit. (Or am I thinking of AB? Meh!)
Nothing to change

Labels containing "Spr"
mb.tchyel   MB242Spr <--kicking this can down the road until canmbs, Provincial Roads

Directional suffixes, EG MB6 MB236_W MB236_N
mb.mb005   MB274 -> MB274_E
mb.mb006   MB236_N -> MB236
mb.mb006   MB236_W -> 67N: compare 2016 vs earlier imagery. Downloaded? Apparently so.
mb.mb023   MB422_N -> MB422

NMPs
mb.tchmai: MB10 +-> MB10_S
mb.mb004: resync MB9
mb.mb006: resync MB101
mb.mb010: resync MB1_W & MB1_E
mb.mb101: resync MB6 & exit 1

Operation Multiplex
Everything looks good

Datacheck
mb.mb039;X809743;;;LABEL_LOOKS_HIDDEN;
Everything else

TCH carbon copies
Winnipeg route labels

[oscar]

SK/MB coords: all but MB57

Nothing to coordinate, other than the TCH routes (which we both coordinated some time ago). Not even the handful of in-dev SK routes reach the SK/MB border. Just set the MB route border points in the right places, and when the remaining SK routes get developed we can just pilfer your MB border point coordinates.

TCH carbon copies

Might be more efficient to save that until just before activation, so any updates before then need not be done in two files.

[yakra]

Nothing to coordinate, other than the TCH routes (which we both coordinated some time ago). Not even in-dev SK routes reach the SK/MB border. Just set the MB route border points in the right places, and when the remaining SK routes get developed we can just pilfer your MB border point coordinates.

Correct. This item is on the list because none of the in-dev SK routes reach the MB border. During my late-night post-GISplunge proofreading sessions, I mostly (probably) blindly replaced SK/MB coords to match the border points on your side -- which don't actually exist. In the interest of being overly anal, I want to go back thru and put the GISplunged coords back in. Rather than just have them be, you know, a very close approximation... So after a fashion, setting the border points in the right places is what I'll be doing.

Might be more efficient to save [TCH carbon copies] until just before activation, so any updates before then need not be done in two files.

Agreed. Hence this item being down near the bottom of the list.

[oscar]

On the final batch of routes in the preview system:

MB 30: MB14 => MB14/332
MB 31: MB3 => MB3/240

These are for consistency with how those points are labeled, respectively, in the MB 14 and MB 3 route files.

There might be similar inconsistencies in these or other route files. I'm flagging those only because they'll affect my own list file when I send in a cumulative update over the weekend.

[yakra]

There are bound to be more inconsistencies.
I actually changed up a few of these labels between my first and final drafts, but my search was by no means exhaustive.
I'd prefer to do the opposite, and err on the side of label brevity & cleanliness: If an intersection includes both a Provincial Trunk Highway (tier 4) and a Provincial Road (tier 5), just include the PTH in the label. Thus on MB14, MB30/332 -> MB30, and on MB3, MB31/240 -> MB31.
I will add a "labels containing '/'" search to the Final Cleanup Checklist. Also, I'd rather automate the search than do an eyeball scan thru every route. I'll be doing several similar "operations".
Any strong objections?

Edit: Only 3 instances to fix:
mb.mb002   MB22/250 -> MB22
mb.mb003   MB31/240 -> MB31
mb.mb014   MB30/332 -> MB30