usaar: Arkansas state highways

[mapcat]

usaar was activated promoted to preview tonight. Ready for peer-review if anyone wants to tackle 787 routes (for now).

Here are some preliminary ideas and observations I made while compiling the set, adapted from the README.md file on GitHub:

Unsigned concurrencies are treated as gaps between separate routes unless the concurrency follows a single route between splits in the same county.

Instead of Alternate, Business, Truck, or Spur banners, auxiliary routes are posted with suffixes (i.e. AR 00 Business = "AR 00B"). Route names for these auxiliary routes follow the same convention (AR00B instead of AR00Bus, with or without city label). Signage for some spurs is "Spur AR 00S": this is treated as a bannerless AR00S (a similar situation exists with at least one truck route).

All 32 AR980 routes have a suffix so that renaming will not be necessary if any change length or are (de)commissioned in the future.

I'm not sure how to deal with AR43 in OK (specifically its westward bend along OK20): AR43 is concurrent with OK20 along state line, but bends west for some distance to be totally within OK, so it probably should have points where it crosses state line, but no corresponding state-line points exist in the OK20 file.

AR75S is signed "Truck AR 75S", but I've avoided using banners anywhere else. It behaves like a truck route (loop starting and ending on AR 75, bypassing a low bridge), but those are typically signed as AR 00T.

AR151 in usaar is longer than the AR151 in usansf and deserves to be free of a suffix, but is correcting this worth the trouble?

AR922-1A is signed AR722-1A on US165, which is the only signage viewable on GMSV.

Hard to believe that we haven't had a new US set since last August. Enjoy!

[Jim]

usaar was activated tonight. Ready for peer-review if anyone wants to tackle 787 routes (for now).

Just to clarify the terminology for everyone, the system was promoted to preview status, not active status.

[rickmastfan67]

US-71 & US-62 are missing the point for AR-612 (I-49 Exit #77).  Thus, those two routes are broken along I-49.

[yakra]

All 32 AR980 routes have a suffix so that renaming will not be necessary if any change length or are (de)commissioned in the future.

This seems appropriate for these airport roads, and I agree with this solution.

I'm not sure how to deal with AR43 in OK (specifically its westward bend along OK20): AR43 is concurrent with OK20 along state line, but bends west for some distance to be totally within OK, so it probably should have points where it crosses state line, but no corresponding state-line points exist in the OK20 file.

OK20 on its own, if we ignore the concurrency, I'd think is comparabe to ME NH153. There's a bit that dips unambiguously into one state, and another bit that follows the state line, that we've got to include in one state or the other (See also: TX US71).
The state line portion is in the Okladot, but not AHTD shapefiles. (But OTOH, it *is* in the Benton county map -- even the clearly-within-OK portion. So, it's... something of importance, I guess?) I'd say then that this portion should be considered within OK, and thus the state line points are already included: OK OK20 AR43_S OK/AR.
Rather than include this section as part of AR AR43Sil, IMO it'd be better off as part of an OK AR43 route.
OTOH, there's no AR43 sign here, just a MO43 sign. Should we just consider AR43 an unsigned route here and leave it out?
On the 3rd hand, it's a bit early for Missouri, innit? It's not like MO43 continues anywhere south of here. Sign-o for AR43, perhaps? How close are we to craIG county?
I don't see any pressing need for shaping point(s) at the dip into OK, even if I did otherwise for AB17 along the SK border.

AR75S is signed "Truck AR 75S", but I've avoided using banners anywhere else. It behaves like a truck route (loop starting and ending on AR 75, bypassing a low bridge), but those are typically signed as AR 00T.

Looks fine to me; I'd have handled this the same way.

AR151 in usaar is longer than the AR151 in usansf and deserves to be free of a suffix, but is correcting this worth the trouble?

Not worth the trouble; we don't wanna break .list files. There's precedent in MO110 vs. MO110Han on the CKC corridor.

AR922-1A is signed AR722-1A on US165, which is the only signage viewable on GMSV.

[froggie]

Unsigned concurrencies are treated as gaps between separate routes unless the concurrency follows a single route between splits in the same county.

Given how ARDOT has done things for decades, unless you have conclusive proof (whether the rare signage or it appears on maps or in route logs) that a concurrency exists, leave it out entirely.

[mapcat]

Unsigned concurrencies are treated as gaps between separate routes unless the concurrency follows a single route between splits in the same county.

Given how ARDOT has done things for decades, unless you have conclusive proof (whether the rare signage or it appears on maps or in route logs) that a concurrency exists, leave it out entirely.

Do that, and the number of routes included would probably increase by 50%.

[mapcat]

US-71 & US-62 are missing the point for AR-612 (I-49 Exit #77).  Thus, those two routes are broken along I-49.

Thanks. Fixed in my copy.

[froggie]

Unsigned concurrencies are treated as gaps between separate routes unless the concurrency follows a single route between splits in the same county.

Given how ARDOT has done things for decades, unless you have conclusive proof (whether the rare signage or it appears on maps or in route logs) that a concurrency exists, leave it out entirely.

Do that, and the number of routes included would probably increase by 50%.

Wouldn't be the first region with lots of route lists.  And since ARDOT doesn't actually consider these concurrencies, it would also be the proper/correct way to do it.

[mapcat]

Wouldn't be the first region with lots of route lists.  And since ARDOT doesn't actually consider these concurrencies, it would also be the proper/correct way to do it.

True, but all they would all be suffixed routes, making it much less user-friendly than usaky or usapr, whose route numbers are almost all unique.

[neroute2]

They're obviously intended to be one continuous route. Except 159. F that route.

edit: obviously a few others are separate routes too.

[mapcat]

They're obviously intended to be one continuous route. Except 159. F that route.

Yeah, that one was my personal unfavorite.

43 makes very little sense as a continuous route. There has to be a story behind it.

[neroute2]

43 makes very little sense as a continuous route. There has to be a story behind it.

The west piece was designated in 1971 (replacing AR 99) to match Missouri, but the east piece was not renumbered. A similar thing happened with AR 37 in 1976 (replacing AR 47).

[neroute2]

Should 130 be truncated to 1B? That's where the official end is, and the supposed overlap is only signed north/westbound (sometimes with to plates above west 130). Eastbound on 130, you get to 1B and 130 signage just ends. This would affect the files for 130, 1BDew (AR130_N -> AR130), and 1 (AR1B/130 -> AR1BDew_S - or AR1BDeW_S?).

1: AR1BDew_N -> AR1BDeW_N?
1: AR243 -> AR243Mar?
1: US64Bus/284 -> US64B/284
1: US64/64Spr -> US64/64S
1: LawRd -> CR407/428?
1/US 49: WhiRd -> CR701/902?
1/US 49: US49Bus* -> US49B*
1/US 49: US49Spr -> US49Y
1/US 49: ClaySt -> CR542?
1McG: US165 is a wye junction; maybe move the point to the center or the south intersection (which has right of way where the two branches meet)
1BDeW: AR130_N -> AR130_W (or just AR130 - see above)
1BVan/364: CR358 -> CR358/362
4: AR4SSpr -> AR4S
5: 5MiRd -> 5MileRd
5: CR46 -> StaCemRd (Staggs Cemetery Road)
5: MocCreRd -> CR227?
5: US62Bus_W -> US62B_W, US62Bus_E -> US62B_E
5: the block of Main Street is one-way northbound, so move US62B_W southwest a block to the main intersection?
5: AR178_E -> AR178_S?
5Lit: BalGate -> BalGateEnt
5Lit: NBJamRd has been realigned east
5Lit: I-30SerRd_E -> I-30_E (just as the west split is I-30/35, not just AR35)
5Lit: add SprRd and BryPkwy (access to I-30; the latter is what OSM and the Goog call Echo Lake Blvd)
5Lit: US70Bus -> US70B

[mapcat]

Thanks for doing the review.

Re 130, I'm inclined to leave it since it is signed in one direction but welcome input from anyone else with an opinion on this. It won't be the only instance where signage and official ends disagree.

Re the names for points at usausb routes, I followed the convention established by yakra when he did Operation Arkansas Cleanup prior to handing the state over to me. Compare New York and Ohio, which use US20Alt point labels for intersections with routes signed US 20A. I didn't use the same convention for points at routes such as AR4B on purpose.

I'll address the others next week when I get some more editing time.

[neroute2]

I disagree with but understand using Bus/Spr etc. But surely at least "49Y" in a shield should be US49Y, not US49Spr?

7: US167Bus_S -> US167(15), US63_N -> US167(16), ChaRd -> US167(17)
7: US79B_N -> US79Bus_N (ouch)
7: AR128_S -> AR128_E, AR128_N -> AR128_W
7/8: CR340 -> CR311/340
7: NavTrl intersects farther south (opposite Fargo Lane)
7: Ar88_W -> AR88_S, AR88_E -> AR88_N?
5 has DeSotoBlvd and 7 has DesBlvd for the same road. I don't know which is correct, since the name isn't signed in public areas.
7: USFS11 -> FR11? (street sign says Forest Rd 11)
7: USFS36110 -> FR86 or Trl86 (street sign says Trail 86, but shield is the same as FR11)
7/16: LurLp -> ToAR123_N? (name is unposted)
7/16: OldAR7 -> CR7120
7: GumSprRd -> CR8675?
7: CR80 -> CR2890
7: AR7SSpr -> AR7SMar
7B: Bro -> BroSt
7SMar: SkiLoopDr -> CR3353