[OSM-talk] NCN refs - consistency

David Earl david at frankieandshadow.com
Thu Jul 12 09:29:55 BST 2007


On 12/07/2007 09:21, Andy Allan wrote:
> On 7/12/07, Martijn van Oosterhout <kleptog at gmail.com> wrote:
>> On 7/11/07, Peter Miller <peter.miller at itoworld.com> wrote:
>>> Sustrans also has regional routes (ncr) but we don't have any in Suffolk to
>>> my knowledge. We do have regional routes, for example 'Suffolk Costal
>>> Route'. I haven't encoded these but proposed using rcn_ref.
>> Can we please *not* invent a new tag for every possible source of
>> routes. In the netherlands that are dozens of organisations that make
>> cycling routes and we're certainly not proposing adding a new tag for
>> each.
>>
>> We need something like:
>>
>> cycle_route=foo;bar;baz
>> bar:ref=A110
>> bar:name=Bloemenroute
>> baz:ref=A111
>> baz:grade=5
> 
> Umm, I don't understand. Isn't baz:ref a new tag? How is that any
> different from what we're proposing with rcn_ref?
> 
> Andy

The point is that, yes they are new tags, but they don't need 
documenting and defining separately for every value of 'baz', because it 
falls into the same definition as ref, name etc., but relates the ref, 
name etc together.

This is more or less what I proposed, but (a) I'd prefer the common part 
to come first because that's more consistent with other uses of colon 
qualified tags, and (b) I think it should be generalised to route, not 
just cycle route; the fact it is a _cycle_ route (and a cycle route of a 
specific kind, e.g. national cycle route) is then determined via a 
further tag. This is because all routes have the problem this scheme is 
trying to solve, not just cycle routes.

Ass someone else said, though this is a little complicated, using a 
variable to relate tags together, ditors can simplify the presentation 
to the user.

In essence it is a way of having an array of [cycle] routes.

David




More information about the talk mailing list