[OSM-talk] highway=cycle&footway

[Andrew Chadwick (email lists)]

David Earl wrote:
> On 23/03/2009 14:18, Richard Mann wrote:
>> OK. So I get fed up because one of these OSM types insists on retagging 
>> something that I think is a cycleway as a bridleway just because it's 
>> got a sign.

Hiya! That would be me.

"Just because it's got a sign" is the best evidence we have to date. But
of course only about the legal status.

>> I don't have time to have an edit war & I get so fed up I'm close to 
>> giving up on OSM entirely, but decide to have one last stab at making 
>> sense of it.

I'm sorry you feel that way. I too try not to participate in edit wars.
Feel free to discuss any changes I make with me any time, and let's
always try to resolve any disagreements.

I don't see the data as "mine". More everyone's. But I do try to make it
accurate to the extent of my ability.

And I think what you've been doing locally is pretty darn impressive
too: cycle networks and links I wouldn't necessarily think of adding. So
kudos for that, and don't think it isn't appreciated. It is.

>> So I float a proposal on talk-gb, for highway to be physical (which 
>> means it must distinguish all major path classes, so I propose 
>> highway=cycle&footway to plug what to me is a big gap), designation to 
>> cover the law, and access to cover routing.
>>  
>> So rather than make an off-hand comment on a different list under a 
>> different subject, why not just say "I like", "I don't like", "I don't 
>> care" or "but..." or something vaguely helpful??????

I don't like. This duplicates what highway=path does, as you've
described it.

  http://wiki.openstreetmap.org/wiki/Tag:highway=path

This has good, widespread usage. This, plus a set of access tags seems
to fit the gap you describe. It's remarkably flexible, by design. Am I
missing something about what you describe?

> If it says bridleway, then surely it is a bridleway, since these tags
> are informed by evidence on the ground, not a statutory database of
> legal rights. How do you _know_ it is a cycleway if the sign says otherwise?

Ah. In this case it says both (... of course. It's never simple).

In this case, Richard's right in that it's an old bridleway still used
by horses for field access. But it's also been half-surfaced nicely for
bicycle use, and has blue low-flying-bicycles signs along it. And a sign
saying "bridleway" and hoofprints. Oh, and nearby riding schools and
horse mounting steps. And lots of foot traffic, plus private motor
access. It's pretty much the definition of shared use in path form.

http://www.openstreetmap.org/browse/way/31871564 if you're interested.

I still say bridleway (which is how OSM ought to write down the official
classification), but perhaps with more "=designated"s than most
bridleways. And proper surface tags (difficult because it's mixed: one
half is tarmacadamed for cycle and foot use, the other is unpaved for
horse use).

> If you're a horse rider, you might well be just a peeved the other way
> by you changing "their" bridleway to a cycleway.

> In the UK bridleways are legally accessible by cyclists, so if you're
> trying to use this in a context of cycling, why not just recognise
> bridleway as something you can cycle on, or change the access tags to
> make this clear. You could also set the surface tag to indicate that it
> is a paved surface and thereofre better suited to cycling than most
> bridleways (if it is, of course).

IMO this is exactly the right way to approach it. The right answer is
always better, more complete data.

> If you're just worried that mapnik *shows* it as a bridleway, then
> surely Mapnik is right to do so in this case if that's what the sign says.

Arguably the cycle layer should render them a little more similarly.
Although its current blue dashed+yellow casing rendering for cyleways
looks close in my eyes to the green dashes+white casing rendering it
currently uses for bridleways.

-- 
Andrew Chadwick