Extended motorway segments along WA-500 to end/begin at NE Falk Road and NE Stapleton Road, instead of in-between. http://www.wsdot.wa.gov/NR/rdonlyres/706FD0DF-CD12-42CD-9FEE-65545EF3F10B/0/Q14.pdf
This is incorrect tagging; the motorway is not yet complete. Recommend highway=trunk, construction=motorway until the entire segment is upgraded, making it consistent with tagging usage in the US for trunks and motorways.
This issue has been referred to talk-us and the data working group.
Are you referring to the entirety of WA-500 as it exists between I-5 and Fourth Plain Blvd? Or just between the Andreson and St Johns junctions?
highway=trunk + construction=motorway isn't correct tagging. That can break routing.
I 5 to Fourth Plain. construction=motorway shouldn't break routing because it's not highway=construction.
Broke Mapquest Open routing when I did something similar on a road in the past.
Because they picked up on the highway 'type' in the construction tag. Anyways Paul, did you know that WSDOT considers what's tagged right now as a 'motorway'? http://www.wsdot.wa.gov/NR/rdonlyres/706FD0DF-CD12-42CD-9FEE-65545EF3F10B/0/Q14.pdf
What WSDOT considers it at what OSM considers it are two different things, the only country where the official classification and the OSM classification are congruent is the UK. Essentially, if it's not interstate standard, it's likely not a motorway. See the United States entry under http://wiki.openstreetmap.org/wiki/Highway:International_equivalence
Re "construction" confusing routers, I'm not seeing that at http://www.openstreetmap.org/directions?engine=mapquest_car&route=52.63936%2C-2.47228%3B52.63906%2C-2.47056 where http://www.openstreetmap.org/way/121769393 has a construction tag. Can anyone find somewhere where it does? I was using http://overpass-turbo.eu/s/beW to look for candidates.
It happened in the past. I guess they might have fixed that bug with their router.
So, Paul, with the "if it's not interstate standard, it's likely not a motorway", are you going to go edit in NYC and change the 'non Interstate standard' segments on some of the Interstates that were grandfather in with I-# at the start of the system? Or maybe change the 'Super-2' segment of I-93 where the 'Old Man of the Mountain' was down to Trunk? Or how about PA I-376 Business Loop (https://www.openstreetmap.org/relation/1014596) which is a 'motorway' for all but three small 'at-grade' intersections in the middle which between them, are tagged as trunk, while everything else is motorway. Would you come into the Pittsburgh area and 'downgrade' it? If you did, I'd fight you tooth and nail over any downgrade along I-376 Business. Locals should hold a little more sway than people from out-of-state when it comes to this stuff. Anyways, WA-500 looks almost exactly like PA I-376 Business Loop, and I support the way it's tagged at this given moment, especially since some locals say there are 'Freeway Entrance' signs at some of the onramps along it.
No, because it's an exceptional situation in the Interstate cases you mention. In the WA 500 instance, 3 at-grade intersections represents almost half of the junctions on the road, making it more consistent with OSM-US practice of tagging surface expressways as a trunk instead of as a motorway. WA 500, as it stands today, has more in common with, say, US 75 north of OK 11, than, say, I 205 where it crosses WA 500. In both cases, US 75 and WA 500 are a mix of at-grade intersections and grade-separated ramps.
"Any freeway anywhere in the United States, whether designated Interstate or otherwise, gets highway=motorway."
WSDOT considers WA 500 from I-5 to 42nd/Falk and from 54th/Stapleton to Fourth Plain/WA 503 as a freeway. These two segments are built to freeway standards -- four lanes, divided, shoulders a lane wide, speed limit of 55. The entrances to these segments are marked with "Freeway Entrance" signs.
Furthermore, determining if a route is a freeway or trunk based on its length and number of interchanges and/or intersections is arbitrary and asinine, undermining the accuracy that OSM strives for and does a disservice for the end user.
I honestly am not sure what the point of questioning the agency's definition of their own highways is. If WSDOT considers the two segments of WA 500 in contention as a freeway, both by the map linked to in thread above and quoted from their email in the other changeset thread, I submit it's very clear what these two segments is supposed to be.
Comparisons to OR 224 aren't fair or even accurate -- OR 224 is by and large a Super-4 expressway, with the only twinned portions being with the interchanges with OR 99E and with SE Lake Rd.
Also, US 26 southeast of Gresham, while intended to tie into the never built Mt Hood Freeway, doesn't really compare to WA 500, as it likewise only has one interchange (OR 212).
Newberg-Dundee Bypass and Sunrise Corridor, once fully complete, will be freeways, but they're not issues at this time because both will only be Super-2s per their Phase 1 construction.
Also, your comment, "It's kind of silly to suggest that WA 500 is of the same character as I 84," makes little sense - both I-84s are long-distance regional interstates with urban sections while WA 500 is purely an urban freeway with an expressway portion, setting aside the Orchards-Camas segment. That would be like downgrading CA 71 to trunk for its whole length because it's of a different character to I-10.
Essentially, countermanding a state DOT's definition of a route doesn't make sense, particularly when that definition matches what exists in the field. Generalizing what a route is on a map that is striving for accuracy is a disservice to the map's integrity.
WA 500 is, by and large, a super 4. I'm not seeing any functional difference between 500 and 224. OSM's definition as it's presently being used nationally with what is a trunk might differ from what a state's definition of what it may be. It's normal for there to be some variation with the official definition of a classification and what OSM has it down as. But if we're going to call WA 500 a motorway, we might as well call fourth Plain, Mill Plain, Saint James and Saint John's motorways as well, because fuck having different classifications, let's just upward creep the whole map.
WA 500 is divided. It has two carriageways.
Super-4's do not. OR 224 is a Super-4.
The only crossover points on WA 500 are from where the concrete divider begins and ends at the left turn lanes for 42nd, and at 54th.
So no, WA 500, in the segments in contention, is NOT a Super-4. It is a fully divided freeway.
However, neither WA 500 or OR 224 are both fully controlled access (ie, grade seperated) and fully divided. It's limited access (a combination of interstate style roadway with surface intersections) and fully divided, which would make it a textbook trunk candidate.
Do not compare WA 500 to OR 224. They are apples to oranges. WA 500 is fully divided from I-5 to WA 503/Fourth Plain; you are incorrect in saying that it isn't. And again, you are attempting to generalize the road into a single category, which is cartographically dishonest for a map that strives to be as accurate as possible.
I'm saying the common usage for motorway has four parts: High speed limit and all three of 1) full grade seperation, 2) multilane, 3) fully divided. Any exceptions should be extremely rare to non-existant (for example, the ranch access gates for ranchland in Utah that have no other frontage on I80 or the gated off fire lanes coming off I 5 in the Siskyous, or the lone dead-end on keystone lake's Diamond Head from US 412, not a major street every mile). Dancing back and forth between the tags breaks consistency when the nature of the road doesn't change enough to warrant a change in classification just for one block. It's a surface expressway, not that lone block of an otherwise residential road on Southwest Birchwood between 87th and Laurelwood in Beaverton.
If it's high speed and two out of three, then it's a trunk.
And the 'nature' of the road is a freeway with an expressway break. It is not a surface expressway.
To use an absurd simile:
A person has a prosthetic limb. They are wanting to clone their missing limb but can't afford it so the prosthetic must do in the meantime.
By your standard, instead of accurately calling them a person with a prosthetic limb, they are a mannequin (I am discounting the fact that mannequins aren't alive for this comparison).
No one in their right mind would make that argument, yet that argument is being made for WA 500. On a map that is designed for as much cartographic accuracy as possible, countermanding the definition of the transportation agency that built and maintains the route (which should be the final say for mapping purposes), and countermanding the mappers who live in the area and have field checked the route.
This accuracy is ultimately for the benefit of the end user, the people who use OSM as an alternative to the other online maps who have become rife with errors. Putting routes under blanket tags is an error.
So what makes the northwest so special that they can just disregard common usage? I'm familiar with the road, and with WSDOT's idiosyncrasies.
If you were familiar with the road, why call it a Super-4 when it's not? Why compare it to OR 224, when such a comparison is disingenuous at best?
And speaking of OR 224, it is incorrectly mapped as a twinned highway when it isn't.
Indeed you're correct on 224. However, in both cases, neither operate completely as motorways for about the same reason. If we were to tag it your way, it'd be switching back and forth between motorway and trunk every other junction (but you're right, it should be converted if they finally took out the K-rail for ease of maintenance like they were talking about doing). You can have trunks with on and off ramps, which is basically what both are. 99E just off 224's another example of not-quite-a-motorway. WA 500's close to it, and I see merit in tagging it as a motorway in the future once the only at-grade intersection with the expressway portion is 4th Plain, but until then, motorway sets the wrong expectation to viewers and routers, and flipflopping back and forth without any change in the nature of the road is disingenuous. What makes entering the strip from east of 54th and west of Falk different from driving in the only segment of this expressway classified correctly? Nothing, you're still having to look for a traffic light before the end of the high speed portion of the road. If 224's undivided now, then the only real difference between 500 and 224 is 500 has a K-rail in the center of it.
To wade in here a bit, I don't think the North American usage of the "highway=motorway" tag has ever required that a road have freeway characteristics from end to end, merely that the segments tagged as motorway have those characteristics for a reasonable length. I really don't think an appeal to "common usage" flies here.
In WA 500's case, it appears the freeway is interrupted by a segment with two at-grades for maybe 15% of its length from I-5 to Fourth Plain Blvd. This seems somewhat analogous to the case of US 71 in Kansas City, where NIMBY opposition has left a section of the route without full access control. How is it tagged? See http://www.openstreetmap.org/#map=14/39.0107/-94.5474
That seems to be an equally oddly tagged trunk; I'd probably have gone with trunk from Swope Pky to I 435 so there's at least an equal interchange at either end of the segment. These things get backed up more than motorways more or less inherent to their design compared to motorways (if you except extremely rare once in a million cases like the Interstate Drawbridge or the Peace Arch Park border station, for which you have literally miles of warning for, rather than 800-odd feet for a traffic light like you would on a trunk.
From the standpoint of someone who is just as familiar with the area, making the entirety of the road a trunk road when the entrances say "freeway" and the state map says "freeway" is misleading at best. And Mr Johnson, I understand your point that it's inconsistent within the OSM community (though I don't necessarily agree, and I think each road should be marked based on its individual merits), if continuity within OSM is more important than mapping what is actually on the surface of the earth, what's the point to begin with? This is absolutely baffling to me. The state clearly says that the roadway in question is freeway. Hell, if you read the wiki page for motorways, it says "[f]or roads that do not meet the above requirements ***consider*** using highway=trunk". We have considered it, but based on other facts present (the official state map and road signs), highway=trunk would be incorrect.
Well, let's look at it from the root of what the trunk and motorway definitions are, the AASHTO definitions of limited access, partially controlled access, and controlled access. OSM would consider the former two as a trunk, and the latter as motorway. AASHTO's definition would consider WA 500 from 5 to 4th Plain as partially controlled access.
Again, isn't OSM's mission to provide the most accurate maps online through the use of open collaboration? Rigidly adhering to definitions that don't fit every situation compromises cartographic accuracy and disservices the end user. It's bad enough that OSM uses the Mercator projection (and note this is only a critique of the distortions inherent to Mercator, not a social commentary of the projection); why make things worse by stubbornly holding to rules that don't exactly fit situations at hand?
Yes, that is the point, Bickendan, which is why I'm rejecting WSDOT's assertation and substituting in one that is the most consistent in accurately describing this object. One thing we do need to watch out for is an upward creep in tag classification as well or the higher classifications lose meaning. When there's a concise way to accurately describe that, why not do that instead of trying to pass it off as a larger road because a state agency did? What WSDOT has it down as on paper and how it functions in actual operation are two different things.
But this is the point you're not considering: We are classifying how the road functions in actual operation, from in the field observation!
Evidently not, motorways are controlled access, divided exclusively.
From I-5 to 42nd, is WA 500 divided? Is it controlled access?
From 54th to Fourth Plain, is it divided? Is it controlled access?
And, from 42nd to 54th, is it divided? Is it controlled access?
The only real difference between trunk and motorway in OSM classification is that trunks might be controlled access but lack a median, or are divided but lack a full control along a segment that people generally consider as contigous. Certainly that is the case with WA 500, as just saying SR 500 or hearing it in a traffic report on KATU or KEX immediately evokes the entire partially-controlled expressway as a single unit.
Also, I'm going to have to add additional justification to revert this changeset and the related ones along 500 as this also seems to have reversed around 12-15 hours of hard work removing lane tagging I spent a lot of effort on.
Paul: It was *your own edits* that deleted the lane counts. Reverting this changeset wouldn't fix that. See http://www.openstreetmap.org/way/45846830/history and http://www.openstreetmap.org/changeset/32790788
I don't think this particular changeset had any lane modifications, but my original changeset which changed the roadway from trunk to motorway did include quite a few merged ways on my part (I was not aware how much detail there was in regards to lanes). I did not intend to ruin any detail that you had implemented, Mr Johnson, and despite the fact that I believe the segments as drawn should remain as motorway, changeset's #33669446 and #33587233 should be reverted. That said, the segments in question should still be changed to motorway.
Sorry for coming in a little bit hot. Let's go ahead and get that reverted and work to coming to some agreement on classification. Can we come to an agreement that most people in the region view WA 500 west of Fourth Plain as a single unit?
Who exactly reverts the changeset? I would do it, if I knew how to without potentially exacerbating things.
We could ask somebody in the DWG to do the revert, because of being a neutral party.
Yeah, let's get the DWG to handle the revert since there's apparently more than one changeset involved and this is hella-messy.
OpenStreetMap is a map of the world, created by people like you and free to use under an open license.
Hosting is supported by UCL, Fastly, Bytemark Hosting, and other partners.