canbc (British Columbia Provincial Highways)

[mapcat]

Still lots of opportunities to delete unnecessary named waypoints from these routes...

But in the interest of moving this set along, I'll stop waiting for a response and post what I've got. Besides what's written below, and as Oscar has noted, most of these obviously still deserve a few more points in the remote areas. I suggested a few at some of the more logical places where signs were legible.

BC3: NicRivPP -> MtHopeFSR; BraBend -> SumRivRd; ECManPP_W -> CayFlaPP & off; PasValRd -> UppPasRivFSR; BroRockPP off; StePP off; AshRd -> AshRivRd & off; 7Ave -> 7Ave_E; CouAve -> CouRd; 102Ave -> StrCreRd; 62Ave off; CanRd -> MouBalRd; JohCrePP off; ThoRd off; McCarCreRd off; SilSt -> CopSt_N; KloSt off 1 block; FisCreRd -> ThiMouTr; PheRd -> PhoRd; DonDr off; BouDr -> BouRd; NurRd off; LakDr -> WLakeDr; BroRd -> ELakeDr; NamGrePP off (and unneeded if correctly located); 10thAve -> 10Ave for consistency? ; same for 33rdAve; EriSt -> EriRd; BenAve -> KitRd; eliminate KitRd (too many Kitchen and Kitchener Roads in the vicinity); KidCreRd -> KidCreFSR; ElkoGraRd -> BateAve; 8Ave -> 9Ave; SparDr -> PineAve; CorRd off

BC5: 202 off; 228 off; 231 off; JulCreRd -> 238 & off; 256 off; 355 off; 374A -> BatSt; ChiLouWy -> ChiLouWay; McLureFerRd -> McLureFryRd for consistency? ; AgaBayRd off (OSM shows realignment); PriRd ? no sign there, no name in OSM; BouMouRd off; ChuChauCPP ? nothing there; NThoRivPP off; SwaRd off; CleVilRd -> CleVilRd_N; MesStaRd -> DiaDr; FinnCrePP ? nothing there; BlueLakeFSR -> BlueRivArptRd; MainSt_S -> WhiFillRd; MainSt_N -> LosRd; TetJauCacRd -> BlaRd; JacFlaPP off

BC19: Fry_S -> DukePtFry; 5 -> DukePtIndPk (exit is unnumbered); 8 -> 7; 21 off; BayRd -> NorBayRd or NWBayRd (goes to actual Northwest Bay); + wpt @ McNairLakeFSR (per imagery; OSM shows it as ElkBayFSR); OyerRd -> SatRd; MainRd off; + wpt @ CluCamRd (to Cluxewe Resort); InlIslHwy -> DouSt; Fry_N -> BearCoveFry

BC20: No imagery exists west of PypLakeRd, so I only had OSM & the LKI to work with there. KRd -> KopRd; Fir ? nothing there; move PunSt to PunArptRd (more important and signed); + wpt @ BullCynPP; HanCanCrkRd -> BigCrkRd; TieRd -> TleRd; MelCrkRd -> ChiMelRd

BC37: OolAve off; LakLakePP off; OldLakLakeDr off; TerAir -> NeuWay; KruRd off; ClaSt off; SmiPlRd -> SmiPl; CedFryRd -> CedRd_N; CedRd -> KitWRd_W; KitRd _> KitNRd_N & off; 1Ave & 4Ave not signed that way (LKI calls both Kitwancool Access N & S, which may be what the blurry signs say); MeiLakePP -> MezLakePP; NinPP off; BobQuiLakeAir off; KinLakePP off; SitRivPP -> StiRivPP

BC99: 2A & 2B -> 2; 80St -> 23; 41 off; OakSt -> OakSt_N; GraSt -> GraSt_S; eliminate Nel/SmiSt or pick one (wpt seems completely unnecessary); BC1A/99A -> GeoSt_E; 4(BC1) off; eliminate SeaSkyHwy_S since it's only BC1's exit (past 1/99 split); SeaSkyHwy_N -> HorBayDr; TidWay -> KelGroWay; CryFalRd -> BruRd; PorRd_S -> PorRd; combine FurCreRd & WatDr (2 halves of same interchange); MainSt -> CopDr; StaChiPP off; + wpt @ CenWay missing (GSJ to Garibaldi Ests); MarRd -> MamRd; + wpt @ ChaCreFSR & RetLakeRd (both near Daisy Lake); BraFalPP off; + wpt @ CalValRd (Whistler Olympic Park access); WesRd -> AlpLakeRd; TaylWay -> TayWay; EagDr -> WhiCayDr; BlaWay -> SprGroWay; + wpt @ ScoValRd (near Shadow Lake Interpretive Forest); PorRd_N -> VineRd; OldPenFarmRd -> PenFarmRd; PemDarRd -> MainSt; + wpt @ InShuFSR (by Lillooet Lake); JofLakePP -> JofLakePP & off; DufLakePP off; FraRd -> TexCreRd; + wpt @ FouValRd (reservation N of Lillooet); PavResRd -> PavCliRd; MarCanPP off (& Cyn?); BC97 off

Still to do: BC1, BC16, and BC97.

[oscar]

Just some quick notes on BC5, from a quick skim:

For the two provincial park points with "nothing there", I've noticed in other files similar points, just placed in the vicinity of those parks rather than at an intersection. Unless I could find an intersection road where you can drive into the park or park at a trailhead, picnic area, etc. where travelers might leave the road or turn back the way they came, I've deleted those points, figuring that some parks are intended to be accessible only by canoe or other non-motorized means even though they are adjacent to the highway. For FinnCrePP, move and rename to FinCreFSR, for which OSM provides name and coordinates, and Google Maps indicates it connects to a parking lot that might be how park visitors enter on foot or by snowmobile. As partial as I am to waypoints for provincial parks, it's OK to omit or change ones like these.

For PriRd, OSM provides coordinates but not the name, and Google Maps provides the name. I would consider that adequate to put the PriRd name on the waypoint, even if there's no imagery, and especially (but not only) if the LKI identifies an intersecting road there called Price Road.

[mapcat]

For FinnCrePP, move and rename to FinCreFSR, for which OSM provides name and coordinates, and Google Maps indicates it connects to a parking lot that might be how park visitors enter on foot or by snowmobile. As partial as I am to waypoints for provincial parks, it's OK to omit or change ones like these.

Totally agree that there need to be waypoints for all provincial/national parks with highway access, preferably at an entrance marked with the park name. But since the name of the road is actually Finn Creek FSR, and not something like Finnagan, can't it remain FinnCreFSR?

For PriRd, OSM provides coordinates but not the name, and Google Maps provides the name. I would consider that adequate to put the PriRd name on the waypoint, even if there's no imagery, and especially (but not only) if the LKI identifies an intersecting road there called Price Road.

I saw the Price Rd name in OSM too, but it was attached to a different road, not the one intersecting BC5. Although sometimes they're more accurate than OSM, Google Maps isn't immune from errors, so I'm wary of using its names for waypoints for the legal reasons we've discussed before. How badly is a waypoint needed there? Just north of there, where the road turns to the northwest, it intersects Price Rd (visible in GMSV), so IMO that would be a better place to locate this point, assuming it's of any benefit (there are other named waypoints nearby).

[oscar]

Totally agree that there need to be waypoints for all provincial/national parks with highway access, preferably at an entrance marked with the park name. But since the name of the road is actually Finn Creek FSR, and not something like Finnagan, can't it remain FinnCreFSR?

Yes. My bad.

For PriRd, OSM provides coordinates but not the name, and Google Maps provides the name. I would consider that adequate to put the PriRd name on the waypoint, even if there's no imagery, and especially (but not only) if the LKI identifies an intersecting road there called Price Road.

I saw the Price Rd name in OSM too, but it was attached to a different road, not the one intersecting BC5. Although sometimes they're more accurate than OSM, Google Maps isn't immune from errors, so I'm wary of using its names for waypoints for the legal reasons we've discussed before. How badly is a waypoint needed there? Just north of there, where the road turns to the northwest, it intersects Price Rd (visible in GMSV), so IMO that would be a better place to locate this point, assuming it's of any benefit (there are other named waypoints nearby).

Seems reasonable to me, if we can live without a PriRd point. But ISTM that the "legal reasons" against taking coordinates from Google Maps (specifically, GM's "no tracing" rule) don't necessarily apply to things like intersecting road names.

[Jim]

I've always believed that we are safe finding road names from GM, but since harvesting coordinates was more questionable, we avoid that.

[yakra]

ISTM that the "legal reasons" against taking coordinates from Google Maps (specifically, GM's "no tracing" rule) don't necessarily apply to things like intersecting road names.

I've always believed that we are safe finding road names from GM, but since harvesting coordinates was more questionable, we avoid that.

Yes, this has always been my approach.

[oscar]

One note on YT BC97Con (which is being modified by a pull request right now, this can be addressed later):

CoalRivSprTP => (shaping point)

GMSV confirms the only thing at that location is a pullout with a "Welcome to the Yukon" sign. The latest Milepost guide to the Alaska Highway and other Arctic roads doesn't mention any access from there or anywhere else to Coal River Springs TP. The park's website confirms that there is no road access, there's no good way for visitors to enter the park (best way, for the really determined, is by canoe or helicopter), and that the park was established to protect geological formations rather than as a destination for visitors. So this looks like another non-intersection point placed in the vicinity of a park, that no park visitor would ever use.

A side issue, which I view as a non-problem, is that the YT/BC border might be misdefined by both Google Maps (which is what is used for the BC 97 and YT1 route files) and OSM/Mapnilk. GM tends to assume that the border follows the 60th parallel, which isn't always right due to surveying errors (as I found with the east end of NT 5, where GM placed about two miles of the highway in AB that actually was north of the signed AB/NT border). OTOH, that "Welcome to the Yukon" sign indicates OSM/Mapnik is wrong to indicate that all of BC97's Contact Creek segment is within BC. What really matters is that the entire Alaska Highway is covered by route files for BC, YT, or AK, and here consistently using the GM border definition is OK for that purpose.

Now that I'm recovered enough from eye surgery to dive into the Milepost, I might have some more comments on BC 97 north of Dawson Creek BC, including other supposed park accesses that aren't.

[si404]

Cool.

Plus having the border at 60.00000 is not really tracing Google Maps.

[Bickendan]

Looking at OSM, there is a bit of discrepancy of the border's location and the fidelity of the vegetation mapping. It appears that the fidelity comes into sharper focus north of 60°, suggesting that the parallel is the border as the level of detail hasn't been increased south of there.

[mapcat]

Here are BC1 and BC97. Oscar told me that he's comfortable with BC16 as-is, and may make some adjustments if he sees any needs, so I left it alone. So this concludes my review.

Both BC1 and BC97 include some lengthy multiplexes, so I have not re-checked segments I already investigated for the concurrent routes. And as before, there are long stretches in need of waypoints, a few of which are suggested below.

BC1: 1? no exit there; 33 off; why aren't 44 & 44A a single point? ; +wpt @ 49; +wpt @ 59; 144 not a named or numbered exit (just access to a power plant); 165 off; OldBC3 -> OldHopeWay; BC7 -> BC7_E; AleBriPP off; HelGateTram off (& is it point-worthy?); BendFerRd -> BosBarRd; BlueLakeRd off; SkiPP off; GolPP off; OreJackRd -> HatCrkRd; BatCreRd -> BatCreFSR; JunBeaPP off; HareRd -> SavAccRd; BucRd -> GarRd; SelRd -> GreRd; 374A -> 375; DalDr & NinPl -> 384; 386 is signed 388 in the opposite direction; 391 -> 390; 397 -> 396; PinRd -> MonPraRd & off (relocated, no exit # yet); ChaFalRd -> ShuCCRd; BenRd -> BerRd & off; VicLakePP nothing there; SumPkwy -> MeaSkyPkwy; MouCrkRd -> MouCrkFSR; WooCrkRd -> WooCrkFSR; TanRivRd -> TanRivFSR; ColForRd -> ColFSR; DonForRd -> DonFSR; DonRd off; BluSchRd -> BlaSchRd; GolViewRd off; GleCrkRd -> GleCrkFSR

BC97: 246Ave -> Rd22; 310Ave -> Rd18; 326Ave -> Rd9; 350Ave -> FaiRd; 386Ave -> SecHillRd & off; 71St -> TucDr; 9Ave -> 9Ave_E; LarAve off; ChaPkwy -> SkaLakeRd_E; EckAve -> EckAve_E; BenHillDr -> BenHillRd & off; BeaAve -> PriAve; TodRd -> ToddRd; DaiRd -> DaiDr; WestRd -> WesRd_S; PinPkwy_S &_N -> PelPkwy_S &_N; WestsRd -> WesRd_N; BarnRd -> BarRd; [BC1 changes listed above]; BigBarLakeRd -> BigBarRd & off; 70MileRd -> BonRd; HorLakRd -> HorLakeRd; ExeRd -> ExeStaRd; ForGroRd -> CanHenRd; ParkDr off; CarDr -> HicRd (or move CarDr west); TimLakRd -> TimLakeRd; CarNatPP -> WriStaRd & off; HorLikRd -> LikRd; OldRd -> MisRd; JacRd -> JacDr; WestcRd -> WesRd; RobRd off; BeaLakRd -> BeaLakeRd & off; HydRd -> QueHydRd; NorStaRd -> NStarRd; MarDr -> MofBriApp; FroSt -> FroSt_S; ThoRd -> VonLieRd; TerBlvd looks more important than nearby ConWay; NorPulRd off; ChiLakRd -> ChiLakeRd; move OldSmiLakRd (wrong name) to SalValRd; CroRivPP off; DaveMusRd -> DavMusFSR & off; KerRd -> KerLakeFSR; CarLakeFSR -> CarpLakeFSR; WhiPtPP off; CampLakeRd -> CarpLakeRd; +wpt @ Powder King Rd (ski area); +wpt @ Hasler Rd (W of Chetwynd); StuRd -> Rd208A & off; Rd237_S -> SemRd; Rd210_S &_N -> Rd221_S & _N; Rd237_N -> SweRd; Rd255For -> Rd255; 72St -> EBypRd & off; MonHwy -> Rd254; OldBC97_Nel -> OldBC97_NelS; Mile306 -> OldBC97_NelN; KleCrePP nothing there; FortHal -> SmiRivPP

[oscar]

Running response to comments on BC1.

no exit 1 -- agreed, and I would remove minor KeiRd point too.

44 & 44A -- the two halves of that interchange are more than 0.5 mile apart, which ordinarily warrants separate waypoints.

add exits 49 and 59 -- OK.

144 -- I'll delete if not needed for shaping (probably), or rename to WahAccRd.

[oscar]

I've gone through the rest of mapcat's changes (agree, except abbreviations for Creek changed from Crk to Cre for consistency with the rest of the route set, much as I prefer Crk). I also found some waypoint additions to remove the remaining distance errors.

I've put in a pull request for my tweaked TCHMai and BC1 files. Also, I've updated the Datacheck FP list to clear the remaining items for BC TCH routes (but not BC 1, BC 16, or other canbc routes, which I'll leave for Si), as well as the one non-distance/angle item for the usaca set.

Once my changes are merged and the site updated, as far as I'm concerned we're good to go with activating canbc.

[Jim]

Once my changes are merged and the site updated, as far as I'm concerned we're good to go with activating canbc.

Would you like me to switch canbc to active on the next update, or would you like one more run at preview status to make sure things look good?

[oscar]

Once my changes are merged and the site updated, as far as I'm concerned we're good to go with activating canbc.

Would you like me to switch canbc to active on the next update, or would you like one more run at preview status to make sure things look good?

I'm OK with that, if you're OK with Si clearing all the canbc false positives post-activation. Since this is the first time I've made changes directly to the datacheckfps.csv file (rather than e-mailing them to you), I'll want to make sure I did that right.

[oscar]

One more thing -- I was going to move usaca from devel to preview, once canbc was activated, so I don't get in the way of that. If you feel like doing both moves at once, go ahead.