The east-west structure as a whole is the Victoria Embankment. It was built for a number of reasons, improved sewerage being the major one. As Richard said, pretty much everyone and everything on there can be said to be “on the Victoria Embankment”.
One of the ways that you linked is an approach to a crossing between a footpath on the Victoria Embankment and the road on the Victoria Embankment. It has a location (“on the Victoria Embankment”) but surely not a name in its own right. Suitable written directions might be “in 200m turn left across the crossing” and perhaps then “cross the Victoria Embankment” and/or “climb up to Waterloo Bridge”.
The other one is a link cycle track between Savoy Street and Savoy Place. I suspect that that won’t have a name in its own right either.
So you guys are saying that if a sidewalk crosses a road on the street then that crossing has no name? ;D
and in the example of the first post
i can indeed see it is not named xD
I mean, it is in the middle of named road of the street, but you won’t name it? ;d
Well, okey. And Richard claims that names can be given, but not always?
Some people will name em. Some people wont. Others will think that crossings are continuous part of a footway and name it, others will remove the name as invalid.
You guys need clear rules on what to name and what not ;D
Just think it through and remember that osm is a global project. Everyone across the world are supposed to use same rules, which is not always that simple, but here i see Canada roads are the same as my country.
Victoria Embankment might be a very special case where people name the footway/cycleway separately from the street, even tho both has the same name. But then i see on the other way of the road sidewalk is also named, sometimes not, spotty rule don’t make a good one
These aren’t sidewalks. And I already linked to the Montreal example where link footways and crossings don’t have names. I never wrote I intend to assign names to every footway.
OK, so, help me understand the logic here:
There is a street in the real world
It consists of two car carriageways, maybe a separate cycle track, and two sidewalks (example)
Separate OSM ways represent the two carriageways, the cycle track, and the two sidewalks
The two carriageways are considered to be part of the street and thus have name=* matching the street name
The two sidewalk ways don’t have a name=* because…
car brain?
they’re not part of the street?
renderers will get upset by the clutter?
per @woodpeck, some parts of OSM are “somewhat shit”, so maybe the fact that there are two carriageways with name=* is also considered somewhat shit, and so we shouldn’t add more shit like name=* on sidewalks?
Maybe we’re waiting until we can move name=* from the carriageways to a street relation?
We actually have a couple of street relations in Toronto, so maybe I can add the sidewalks to the relations, and then delete the name from the car carriageways? Will this be somewhat less shit? Or are we only okay with nameless ways for non-car data?
something else I’m missing?
Please comment.
I actually agree personally, but as you are aware it very much isn’t the tagging norm in most of OSM (example), and it isn’t the norm in Toronto, and the tagging norm is also enforced here by non-Canadian mappers like Telenav… so overall I’m not sure why your corner of the world is relevant.
No I have not tried. (Insert rant about my experience with OSM tools being coloured by osm-carto.) Has anyone else tried? Maybe the Pedestrian Working Group? Or is everyone else just straight-up ignoring the sidewalk ways, like Minh’s coworkers?
At the time, I failed to convince the original mapper to just tag the same wikidata=* value on each of the roadways. If a street needs a relation, I think street would be a fine name to call the relation type, but unfortunately someone got there first with a bizarre idea to put literally every node and way in a city inside one or more of these relations.
They were clients, not coworkers. Another Mapbox customer recently launched with “sidewalk differentiation technology”. I haven’t tried it, but I assume it’s also keeping sidewalks out of the guidance instructions. Like the client I supported, their application doesn’t display a map at all, so the user can’t tell that it’s actually calculating a route down the street. All they know is that it’s announcing street names instead of micromanaging their footsteps as if they’re a sidewalk delivery robot.
The Valhalla option on osm.org is configured with worse behavior: it sometimes traverses the sidewalk, sometimes traverses the street, and even the protected bike lane, rather unpredictably. Apparently the site requests routes with a completely neutral walkway_factor and sidewalk_factor of 1.0, making streets and service roads just as attractive as sidewalks. This gets you the worst of both worlds: overdetailed instructions and missing street names, plus extra turns that you shouldn’t make.
Experimentally, I find 0.25 to be a much more reasonable value for both factors if you actually want to traverse sidewalks, or a much higher value if you don’t. As someone who focuses so much on making things data-driven, I find that Valhalla’s geocoder-like emphasis on tunable factors with mysterious units really drives me batty sometimes, but it is quite flexible.
ya cant say that something is tagged as sidewalk but then say it is not a sidewalk ;d But not telling you what to do obviously, just hinting pitfalls!
as for streets. Streets are areas, not lines, like i explained earlier. Everything in that area is named. Really doesnt matter if it’s for horses, trams or whatever. If it’s a benchat you sit on, then you sit on this street. The bench is a part of it.
now, why do we name only the road? Because in OSM we use simple tools to map the reality. Not because we love cars (But i heard you gotta kind of a problem with car-centric street design, so just please, please, don’t make it a global problem).
So, because road is the central part of the street and it is also a line, this is why we name it. That’s the only reason. We name it for it to display on printed maps, your screen and other generated results.
If it is two carriageways, we name both of those lines, because otherwise we would have to draw a third line in the middle, tag it with the name and with something else as well so it actually renders. But if we would do that, we could actually use the relation i described and so you could add all your sidewalks, cyclopathy, roads and whatever (including buildings that has this street address) and yes, then we wouldnt name the road, coz it would be assigned to the named relation,
but since it’s a draft, we need renderers and routers to actually start using it.
I’m sorry if i didn’t answer all your questions, but right now that quidelines ya ppl propose for naming sidewalks are unprecise and IMO the only reason you do it is to have it used in routers that refuse to use relations (which is our fault, coz we never made official specifications and not even main OSM map use it.
When i ask how to name sidewalks on screenshot of streets intersection, you refuse to do so, so really how ppl are going to know?
I would encourage you to at least try. I will probably take say 10-20 minutes at most per issue, far far less than time needed to even answer questions in this thread.
Thank you for the comments, everyone. It is nice to know that a lot of people care about Canadian mapping!
From the discussion so far, it appears there isn’t a large interest in pedestrian infrastructure among Toronto mappers, and there isn’t an accepted way of associating sidewalk ways with street names for purposes of routing among international mappers.
Consequently I have decided to pause my involvement in cleaning up and mapping pedestrian infrastructure. I wish Pedestrian Working Group, Verkehrswende-Meetup, and router developers best of luck. I hope that in a few years there will be a widely supported standard for pedestrian infrastructure in OSM, but I do have to say I won’t be holding my breath. (Prove me wrong!)
If you’d like an opportunity to put a thumb on that scale, we would definitely be happy to have you (and anyone else!) join in participating in the PWG. It’s a great forum for exactly these types of debates - and I’ll reiterate that this specific topic hasn’t gotten much attention yet.
Interesting. I was wondering, is the PWG a discussion space that requires an account to access? Are there protocols or summaries from the monthly meetings available somewhere?
With a quick search I couldn’t find much mention of this group in the forum, so I got curious. Is the PWG mainly focused on the U.S. context, or does it aim for global applicability? I came across this wiki page and this blog post, but I might have missed something else.
I really appreciate the effort to bring clarity and improve tagging practices for pedestrian-related data. These are just some questions that came to mind right now.
One thing I noticed: the term “Working Group” might create the impression that it’s an official OSMF group, like the DWG or LWG. I found it a bit surprising to see it mentioned in passing in this thread without prior context. (Yeah, i know “working group” was not invented by osm or anything like that )
As said, maybe i really just missed something. All the best!
We interact primarily in our scheduled Zoom meetings (Monthly on the third Tuesday, 12pm-1pm PST). Between those meetings, most discussion happens in a dedicated channel we have in the OSM US Slack.
Yes, we collaboratively maintain a Meeting Minutes-style document here.
Both - most participants so far have been focused on the US, but any schema recommendations we produce should be compatible anywhere.
OSMUS started setting up its own working groups a few years ago for topic areas that a local chapter like OSMUS would be better positioned to facilitate than a global organization. These working groups primarily focus on the U.S., but there are members from abroad who provide much-needed international perspective and connections.
So far, there haven’t been any naming conflicts between the working groups of OSMF and OSMUS. If there ever is one in the future, like if OSMF sets up an Imagery Working Group, I assume we’d start referring to the U.S. working group as such to avoid confusion. It would be a good problem to have!
And now Toronto (and a bit of GTA) has 5800 ways tagged with is_sidepath:of:name, with 590 unique names, amounting to 560 linear kilometres of ways (so about 280 km street centrelines): is_sidepath:of:name | Keys | OpenStreetMap Taginfo Ontario
West side of central Toronto is pretty well covered, and I’m slowly working to connect up the islands:
Along the way I also did a whole bunch of cleanup of pedestrian infrastructure, including splitting out crossing ways (previously sidewalk ways often spanned multiple blocks), updating crossing tags, setting surface tags, and improving alignments which in some parts were 15 years old and not particularly accurate. Thank you for the motivation!
At one point, the numbers in Ontario were higher than street:name ways and unique values in all of United States (the mappers there prefer that key), though they’ve exceeded it now; but because that’s spread out across the states, Toronto remains the most-covered city on the continent.
And routers are beginning to pick up the data. GraphHopper as embedded on osm.org now includes names in walking directions in areas that have the data thanks to this code change; an OSRM PR is merged and awaiting release. As before, there’s lots of room for improvement with walking directions, but at least there’s some names in the directions now.
I’m tempted to follow suite with other Canadian mappers here since it looks like Toronto and now Quebec have settled on an approach. Before I recommend anything for Regina SK, I will reach out to mappers in Saskatoon for thoughts. We’re currently trying to move our sidewalks to geometry, while they seem to have very good geometry established.
)