From OpenStreetMap Wiki
Jump to: navigation, search

Wet Paint becoming a Chalk Line: when does Proposed rise high enough to include in OSM?

I have held off for a long time entering this, but I feel it is time as I read Harry Wood's idea of "compromise." Harry says we might "apply some strict criteria for when a proposal is finalised enough to be regarded as verifiable (even though it still doesn't exist)." YES! This really IS a solution to OSM's use of "proposed."

I consider the question by my Subject/headline above to succinctly capture the idea. In it, "Wet Paint" is when a government body publishes a tangible plan to create/construct something (a new road project, a wider bridge, a bicycle route where one didn't exist before...). It is paint because it is official and promulgated to the public, but it is wet because it doesn't yet exist in the real world. It is, however, a fully-formed idea. It may even be as far as a bid, tangible construction plans, and/or crews working -- the important thing is that it WILL happen. It is not just an idea or a proposal, it is MORE than that. By "Chalk Line" I mean that the essence of this is captured in OSM, by what is usually rendered as a dashed line along the right-of-way of the project. (Something with the tag "proposed").

In our USBRS_WikiProject, much has been said about Proposed. What started out in OSM (circa 2011) as bold and highly inaccurate assertions of Proposed national bicycle routes, have become a highly tuned set of both Actual routes and Proposed routes, thanks to this project's lengthy conversations about "What makes something Proposed, anyway?" The results can be read at that wiki, but in short, it is that as national bicycle routes in the USBRS are asserted by states, state at a time, only when a state has a true "project" to create a USBR, one which is nearing its final step of completion of application to AASHTO (similar to how a state highway might be promoted to an Interstate highway, and by exactly the same organization) is when a route might be considered to be truly "a proposed route." We call this a "high bar standard."

Recently, OSM-US has asked AASHTO for specific permission to enter routing data into OSM which come directly from AASHTO's web site. These route data are in the form of a ballot from state Departments Of Transportation (DOTs): "Here is our route proposal, AASHTO..." and all AASHTO must do is approve it. (This is to assure specific protocols to achieve a national harmony are followed; indeed AASHTO has never disapproved any ballot, as all are widely vetted within the state of origin and promulgated by the DOT). AASHTO has tentatively given OSM this permission, OSM only awaits final paperwork. After this is received, OSM might say that we have a "very high bar standard" with which to enter a route proposal into our database: a ballot from a state DOT to AASHTO. To wit, this route actually IS approved by the People of the state of origin, AASHTO is simply "rubber-stamping" its inclusion into the national network. Hence, OSM might accurately characterize this entity while it is waiting for AASHTO approval as "proposed."

This can be extended to similar sorts of "official approval pending" or "official plans published" (wet paint) becoming some entity (a route, a road...) in OSM which is well-described as Proposed (chalk line) -- even down to the local level. It took a lot of discussion to get a "high bar standard" to emerge in the case of the USBRS, and it feels fortunate to have discovered what seems like a "very high bar standard." Without getting lost in a miasma of details about what the guidelines might be at every single local level, let us discuss this in the instant case (the USBRS and AASHTO having received a state ballot for a USBR as a "very high bar standard"). Thank you. --stevea 23:28, 8 December 2014 (UTC)

Multiple plans can create a mess?

If two people submit plans which are "nearly but not quite the same", then anyone rendering planned=primary will see both possibilities overlaid on the same map.

So for example, if two designers have differing ideas about exactly where a roundabout should be, then your map might have a load of overlapping roundabouts looking like the olympic rings with a tangle of roads attached, and no way to tell what's what.

Surely the tag would need to say which specific plan this object is part of, rather than by "all possible plans". Then you can generate different maps showing each plan. So tags more like "planned:bedford_council_bypass_proposal_A_2008-10-10=primary" than "planned=primary". Ojw 08:25, 12 October 2009 (UTC)

What about using (in that example) operator=bedword_council and plan_ref=A_2008-10-10 if each plan uses separate ways? Alv 09:04, 12 October 2009 (UTC)
I agree there should be subtags for the source. What about source=* ? --Dieterdreist 19:02, 13 October 2009 (UTC)
Separating plans might be done on a relation level, so that several plans can use the same ways. Also a discarded=* can be of interest to eliminate discarded plans. --Skippern 21:52, 13 October 2009 (UTC)

I'm using proposed only where the land is already owned by the government, and thus, barring major replanning, the alignment is definite. --NE2 17:53, 21 January 2010 (UTC)

Yes. If there are multiple competing proposals, then that seems like quite an early stage of planning.
I would suggest that the data doesn't belong in OpenStreetMap when something is at a woolly and indeterminate stage of planning (where there are multiple different plans) You should wait until a proposed development is more settled/finalised. Otherwise we're mapping ideas which are not very close to any kind of reality. It's not verifiable
-- Harry Wood (talk) 14:26, 16 September 2014 (UTC)


Following on from the above point, there's a general verifiability issue with this whole tagging scheme. I think I'm in support of removing it from the rendering because it's problematic

We have a warning on the page which goes some way to explaining things: "Note that generally, OpenStreetMap requires information to be verifiable on the ground. Proposed features that are not visible on the ground should therefore only be mapped in exceptional cases and after a discussion with the relevant OSM community, lest they might be deleted for lack of verifiability"

And I think that's good, but is there some way we can nail this down some more? In general I'm inclined to say that OpenStreetMap is not a database of woolly non-existent ideas. We map the real world. Things which really exist. It seems to me this whole tagging scheme is in breach of that principle.

On the flip-side I can see that major proposed constructions of roads are something people expect to see on maps. I think actually that's born out of what we get with road atlases, and with modern regularly updated digital maps, it becomes less necessary. I guess a compromise would be to apply some strict criteria for when a proposal is finalised enough to be regarded as verifiable (even though it still doesn't exist)

-- Harry Wood (talk) 14:26, 16 September 2014 (UTC)


Hi, I wonder how to set key:bridge for a proposed bridge (road or railway).

Version 1: highway=proposed; proposed=motorway; bridge=yes

Version 2: highway=proposed; proposed=motorway; bridge=proposed

Any opinions? Any experience with rendering? (Please also see:

--NachtSpazz 21:33, 15 January 2013 (UTC)

render proposed railways

It says "railway=proposed will probably never render on Standard maps". Why not? A railway or subway that is being planned is very interesting data since it will change transport in that region a lot! It would really apreciate it if railway=proposed was rendered.--Geogast 14:05, 5 February 2011 (UTC)

Mapnik display

Currently highway=proposed displays, no matter what the value of proposed=*, as if it were a motorway under construction, but with turquoise instead of the darker motorway blue. It seems that it would be better to have essentially a thinner version of construction, with the normal symbol thinned to 1/2 or 1/3 of its width and then dashed. --NE2 21:26, 21 January 2010 (UTC)

I've added an entry to the bugtracker (wishlist). My suggestion is a rendering like the attached file. Maybe solid instead of dashed lines. --Nibbler 22:04, 6 March 2010 (UTC)

key:proposed for a route or a route member?

Sometimes cycle route lack some cycleway part wich, but you can use the main road to link them. This part without cycleway are not officially part of the cycle route, but will be soon, when the roadwork will be done. Tag this cycler oute part as proposed could help to understand a cycle network under construction.