On Wed, Jun 29, 2011 at 2:35 PM, Nathan Edgars II <nerou...@gmail.com> wrote: > On 6/29/2011 3:28 PM, Josh Doe wrote: >> >> On Wed, Jun 29, 2011 at 3:06 PM, Nathan Edgars II <nerou...@gmail.com >> <mailto:nerou...@gmail.com>> wrote: >> >> On 6/29/2011 2:49 PM, Nathan Mills wrote: >> >> It also avoids the inevitable "which way is >> forward and which is backward" question. >> >> Forward is the direction of the way. If a way carries both >> directions of the route, it gets no role (as with directional roles). >> >> >> I'm a little slow here; forward means the route follows the direction of >> the way (order of nodes), so for dual carriageways if the ways are in: >> * opposite directions: they would both have oneway=yes and both use the >> forward role? >> * same direction: one would have oneway=yes and the forward role, the >> other with oneway=-1 and the backward role? I find it a little >> confusing... > > Yes. This is the standard for bus and bike routes, as well as highway routes > in most countries. JOSM makes it easy to sort a relation this way.
For bike/bus routes that makes sense since they may go against the directionality of the way. For highway routes this doesn't seem to make sense and as Josh pointed out is just duplicating oneway information whereas the signed direction of the highway provides new information. I mentioned something about cardinal direction relation roles on IRC last night and I think it was RichardF thought they were silly because he had no concept of a "north/south" vs "east/west" highway. I guess this is yet another thing that is unique to US highways. I suspect most JOSM developers are the same way so that ticket just doesn't register as relevant to them. I'm guessing a patch might be accepted though :) Toby _______________________________________________ Talk-us mailing list Talk-us@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-us