[OSM-talk] How to tag lanes, not ways, was: Deprecating the use of Tag:highway=stop in favour of Key:stop

[Anthony]

On Sun, Aug 30, 2009 at 10:27 AM, John Smith <deltafoxtrot256 at gmail.com>wrote:

> Why is a relation the best solution?


What solution is better?  Your lane-based solution doesn't work if there is
only one lane with bi-directional traffic.  The solution of adding a node
and a direction would be second best, but I think it's clunky adding
multiple extra nodes instead of one relation.

Regardless if it's common sense or not to slow/stop it's a legal
> restriction, however it only applies to one direction similar to a
> traffic calming device that only applies to one direction, however in
> both case it doesn't apply to the entire way, it just applies to a
> single lane of the way.
>
> Just like there is motorway lanes that have different maxspeeds
> depending what lane you are in, it would be pointless to do this in a
> relation, but being able to tag lanes with individual restrictions it
> would solve a number of issues, not just stop signs.


I don't see the practical application.  We're a long way from having our
cars drive for us using micro-level OSM data and some advanced positioning
system.  If you want to set up a system that can handle that kind of detail,
be my guest.  But it's not needed for something trivial like stop signs and
maximum speeds.


> > Whoever provided Google with their data split the way in the example I
> > provided, and does so in pretty much any situation where you need to be
> in a
> > particular lane in order to get where you want to go.  In my opinion
> paint
> > which disallows lane changes should be equivalent to a physical gore area
> as
> > far as mapping applications are concerned (with allowances for
> > approximations in micro areas and lenience when everyone is going
> straight
> > and it doesn't really matter).
>
> At this stage only physical barriers are drawn as separate ways


Are you sure that's always true?  I'm certainly not going to adhere to that
restriction.  If you have a single way, there is an assumption that you can
travel freely between the lanes.  If I have to get into the right lane 100
meters before the intersection, I want the way split at that point, not at
the intersection, regardless if the barrier is concrete or paint.  I suspect
that's the way things are already done at least some of the time.  It's
common sense, really.


> , that
> isn't really the point of this thread, the point is the best way to
> tag nodes or other restrictions on a per lane basis.


I'm not against it.  I just think it's a lot of work for very little
benefit.  And there are probably better solutions.  If you really want to
map every single lane of a road, why not make the lane the base unit and use
relations or some other method to tie the lanes together in areas where lane
changing is allowed (lane changing allowed between way X, Y, and Z from node
A to node B)?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20090830/90944acf/attachment.html>