OpenStreetMap

[agentsteel53]

Check Google for the correct alignments.

You do know that's a no-no for OSM editing, right?

As an alternative to Google, information on numbered routes can be sourced from USGS topo maps – specifically, the older hand-drawn ones, because the new all-digital maps appear to use commercial road data.  The main drawback here is the route information could be years or decades out of date.  All USGS topo maps are available for download here: http://usgs01.srv.mst.edu/MLAD4/?store_url=http://store.usgs.gov/b2c_usgs/usgs/z_lat_lon_matnr_list.do?b2cb2b=b2c&store_target=_top

Interestingly, now that USGS maps are being crowd-sourced for selected types of data, Google StreetView is now one of the recommended methods for field checking certain facts.

this seems like one giant circle-jerk.  add in the Wiki world's general ban on primary research, and you are left with no options.

[vtk]

Check Google for the correct alignments.

You do know that's a no-no for OSM editing, right?

As an alternative to Google, information on numbered routes can be sourced from USGS topo maps – specifically, the older hand-drawn ones, because the new all-digital maps appear to use commercial road data.  The main drawback here is the route information could be years or decades out of date.  All USGS topo maps are available for download here: http://usgs01.srv.mst.edu/MLAD4/?store_url=http://store.usgs.gov/b2c_usgs/usgs/z_lat_lon_matnr_list.do?b2cb2b=b2c&store_target=_top

Interestingly, now that USGS maps are being crowd-sourced for selected types of data, Google StreetView is now one of the recommended methods for field checking certain facts.

this seems like one giant circle-jerk.  add in the Wiki world's general ban on primary research, and you are left with no options.

Fortunately, OSM embraces original research.

[Roadsguy]

I noticed that multiplexes/concurrencies (i.e. ref with multiple values, e.g. I-99;US 220) are now rendered in shields. Previously, there was a character limit on how long the rendered "shield" would be. It seems to have gotten larger, as some multiplexes are showing (like I-78/US 22, I-70/I-76, etc.), but others aren't. I can't find any shown triple multiplex except this one.

Is this a bug? I like it, because there are a lot of long multiplexes like the ones mentioned that don't have any shield at all. A way to render the shields separately should be implemented, though. Is such a feature in the works, or is it not possible? EDIT: Apparently not only is a relation-based shield rendering thing apparently in the works, but also shield images! It's clearly not implemented yet, though...

[rickmastfan67]

They just recently upped the number of 'characters' to be rendered in shields up to 11.  Now stuff in the ref tag like 'US 19 Truck' and 'US 19;PA 51' can now be rendered.
https://www.openstreetmap.org/#map=13/40.4661/-80.0327

There is also in the plans on adding multiple lines for refs via the ';' in tags.  See here: https://github.com/gravitystorm/openstreetmap-carto/pull/750

[rickmastfan67]

We now have routing on the main OSM site!!!!

https://blog.openstreetmap.org/2015/02/16/routing-on-openstreetmap-org/

[Bickendan]

Southbound I-295 in NJ has an odd 'fold'. http://www.openstreetmap.org/directions#map=16/39.6972/-75.4680&layers=Q

[rickmastfan67]

Southbound I-295 in NJ has an odd 'fold'. http://www.openstreetmap.org/directions#map=16/39.6972/-75.4680&layers=Q

Somebody has already fixed it.  MapQuest just hasn't updated yet.

[SD Mapman]

Southbound I-295 in NJ has an odd 'fold'. http://www.openstreetmap.org/directions#map=16/39.6972/-75.4680&layers=Q

Somebody has already fixed it.  MapQuest just hasn't updated yet.

I got it after Bickendan's post... I don't know why it was there.

[rickmastfan67]

Southbound I-295 in NJ has an odd 'fold'. http://www.openstreetmap.org/directions#map=16/39.6972/-75.4680&layers=Q

Somebody has already fixed it.  MapQuest just hasn't updated yet.

I got it after Bickendan's post... I don't know why it was there.

It seems that it was a newbie who did it with P2 per the history on that node for the bridge.

[Bickendan]

While I like the idea of US 66 being resurrected in California, can we get Hist banners up on the shield?
http://www.openstreetmap.org/#map=14/34.1394/-118.1157&layers=Q

[rickmastfan67]

While I like the idea of US 66 being resurrected in California, can we get Hist banners up on the shield?
http://www.openstreetmap.org/#map=14/34.1394/-118.1157&layers=Q

Send a tweet about that to '@MapQuestTech' and link them to the way in question. Then they might add it to their roadmap to fix.

[Bickendan]

Bumping because I'm not starting a new topic.

That moment when motorway vs trunk distinctions get overly pedantic: https://www.openstreetmap.org/changeset/32230005#map=10/45.5578/-122.4845

Briefly: I changed WA 500 from trunk to motorway between I-5 and WA 503/Fourth Plain Blvd, save for the segment between 42nd and 54th. It got challenged, and later reverted to straight up trunk. My latest comment ain't happy about it.

[rickmastfan67]

Maybe a compromise by adding the 'motorroad=yes' tag?

[Bickendan]

Maybe a compromise by adding the 'motorroad=yes' tag?

Wouldn't that technically apply to any and all roads motor vehicles drive on?

[rickmastfan67]

Maybe a compromise by adding the 'motorroad=yes' tag?

Wouldn't that technically apply to any and all roads motor vehicles drive on?

https://wiki.openstreetmap.org/wiki/Key:motorroad

[NE2]

It's Paul Johnson. He's wrong much of the time, but don't dare to disagree with him or you'll get banned.

Maybe a compromise by adding the 'motorroad=yes' tag?

Wouldn't that technically apply to any and all roads motor vehicles drive on?

https://wiki.openstreetmap.org/wiki/Key:motorroad

"The motorroad tag is used to describe highways that have motorway-like access restrictions (i.e. cyclists and pedestrians banned) but that are not a motorway." SR 500 is open to bikes: http://www.wsdot.wa.gov/bike/closed.htm#Clark

[Bickendan]

It's Paul Johnson. He's wrong much of the time, but don't dare to disagree with him or you'll get banned.

Maybe a compromise by adding the 'motorroad=yes' tag?

Wouldn't that technically apply to any and all roads motor vehicles drive on?

https://wiki.openstreetmap.org/wiki/Key:motorroad

"The motorroad tag is used to describe highways that have motorway-like access restrictions (i.e. cyclists and pedestrians banned) but that are not a motorway." SR 500 is open to bikes: http://www.wsdot.wa.gov/bike/closed.htm#Clark

That restriction also doesn't apply to non-urban interstates.

[lordsutch]

Honestly, I can see both sides of the argument. I'm not familiar enough with WSDOT practices to wade in, but to me unless WSDOT has an official map designating it as a freeway or there's some distinguishing field signage between freeways and non-freeways in Washington and you're using that as the basis for what's motorway and what isn't (as opposed to more loosey-goosey stuff like number of grade separations versus at-grade crossings), it's not worth spending much capital on.

And an edit war is not a smart plan anyway. Refer it to DWG if you really, really want to stand on principle, bearing in mind that DWG's interests and those of pedantic roadgeekery aren't likely to coincide (after all, there are whole countries were many fully-grade separated highways are tagged highway=trunk).

[vdeane]

I'd call a road like WA 500 a freeway except between those two at-grades.  Just because it's short doesn't mean the whole road should be treated as one.  Marking the freeway sections as such paints a much more accurate picture of what it's like to drive on that road at any given point.  Looks like OSM is adopting the same methodologies for labeling roads as freeway/surface streets that drove me away from MapQuest.  So much for accuracy...

[SD Mapman]

So if any at-grade makes a road a trunk highway, does that mean US 36 in St. Joseph should be trunk too, due to the two at-grade ramps? 

https://www.openstreetmap.org/#map=17/39.74912/-94.85761

[jakeroot]

...unless WSDOT has an official map designating it as a freeway or there's some distinguishing field signage between freeways and non-freeways in Washington and you're using that as the basis for what's motorway and what isn't (as opposed to more loosey-goosey stuff like number of grade separations versus at-grade crossings), it's not worth spending much capital on.

A couple of points:

Like California, Washington uses "Freeway Entrance" signs at its on-ramps. Along WA-500, here's what each interchange has posted:

St Johns Blvd: "HIGHWAY ENTRANCE"
Andreson Road: "FREEWAY ENTRANCE"
Thurston Way: "FREEWAY ENTRANCE"
Gher Road towards east: "HIGHWAY ENTRANCE"
Gher Road towards west: "FREEWAY ENTRANCE"

Directly east of the Andreson Road junction is an at-grade signal...should be "Highway Entrance" in that direction (to be, like Paul, overly pedantic). Directly east of St Johns Blvd is all freeway, so it should read "Freeway Entrance" in that direction...no "buts" about that one.

Additionally, WSDOT does produce online maps. Here's a snippet of the area around Vancouver. Washington considers parts to be freeway, other parts to be regular highway:

[vtk]

Seems to me a reasonable convention is, a stretch of expressway can be considered a freeway if it is five miles long, or has three interchanges, between at-grade intersections. Looks like you have two segments that marginally qualify. Are there other mappers in that area who can weigh in?

I also agree with adding motorroad=yes, because this is clearly an expressway, and that's the best tag we have to distinguish expressways from important conventional roads that are also tagged trunk. If bikes are indeed allowed, then bicycle=yes is called for. I would say bikes on a freeway or expressway is an oddity in this country, so it's reasonable to have expressway tags that seem to imply bicycle=no countered with an explicit bicycle=yes in this situation.

[Bickendan]

Using a minimum (and arbitrary) length to determine whether a segment is a freeway or not is not a good rule. The rubric should always be: How is it built in the field? How does it function? If it's built like a freeway, acts like a freeway, it's a freeway. If an expressway portion interrupts that, it should be noted. If OSM strives to be the most accurate map around, considerations like functions of junctions to length muddy up the water and dilute the accuracy. At a glance, and this comes from a cartographic perspective as well as an end-user's, if a route changes from one road type to another, I want to see that. A route switching from freeway to expressway can be important in routing decisions. Furthermore, on a more fundamental level, marking a route that changes strictly as trunk is completely disingenuous because trunks are being used for completely grade separated routes, expressways and pure at-grade routes. That is not helpful for the end-user, nor is it accurate.
Around Portland: US 30 along St Helens Rd and Yeon Ave is a trunk. OR 99E along McGloughlin Blvd in SE Portland and OR 224 from OR 99E to I-205 are trunks. US 26 between the Ross Island Maze and SE 26th Ave is a trunk. N Greeley Ave from I-5 to Going St is a trunk; Going from Greeley to I-5 is a trunk. OR 99W between Newberg and Tigard is a trunk. OR 213 between Beavercreek Rd and I-205 is a trunk. US 26 between Powell and OR 212 is a trunk. WA 500 between I-5 and WA 503 is a trunk. Padden Pkwy is a trunk for its length.
Which of these are freeways, for any of their length? If you've never driven these, how do you know? I've driven each one of these, and they are not equal to each other, yet according to OSM standards, they are, save for the segment tags that aren't visible on the map itself.

If it were up to me, and this isn't something I'd want to even try to implement given how established the OSM standards are, I'd have any and all freeway segments marked as such (motorway, for simplicity). All non-freeway numbered highways marked (trunk). Non-numbered, non-highway arteries (primary). Then secondaries/collectors (secondary), and everything else, with non-paved in their own field.

[empirestate]

The trunk designation is still problematic all over the U.S. In the New York area, for one thing, it seems to be used to mean "parkway" as much as anything else. Just looking at the map closest to my home, here's what's labeled as "trunk":

-Bear Mountain Parkway: all of it, both segments. Those parkways in the greater region that are full freeways tend to be shown as such, but those with intersections or just lower design standards, such as the Bear Mountain, tend to show as trunk.

-US 6, east of the Taconic Parkway: probably the closest to the intention behind the trunk designation. It's a high-grade surface highway, almost expressway-like in spots. But there are other parts of the route, and other routes in the area, that seem just as worthy of the trunk designation that don't have it.

-US 6/202 approach to the Bear Mountain Bridge: it's a main route, to be sure, but physically isn't anything like US 6 to the east. Narrow and windy, almost parkway-like in some ways.

And that's pretty much it.

[lordsutch]

The problem at its base is attempting to fit a schema designed for a particular approach to highway designation (that of the UK and aped by Ireland and, to a lesser extent, some continental European countries) beyond that context.

That said the scheme as designed maps reasonably well to the North American use of functional classification (but poorly to route designation). But ultimately it also requires judgment calls, and a willingness to accept that your opinion of how to interpret the guidance won't always carry the day and that disagreement doesn't require being disagreeable (even if you are right and the rest of the project is wrong).

In this specific case: if the state built the road as a freeway and it is signed as such, as Washington does, it is unambiguously highway=motorway and should be tagged as such. Length has nothing to do with it. That's been the tagging rule as long as I remember, and it is worth challenging (again, not doing so in a disagreeable way).