On Tue, Sep 1, 2009 at 7:39 AM, Anthony<o...@inbox.org> wrote: > On Mon, Aug 31, 2009 at 3:58 PM, Roy Wallace <waldo000...@gmail.com> wrote: >> >> What do you mean by "it allows nodes to be attached to individual >> lanes"? Do you mean you don't think a lane should be represented by a >> series of nodes? > > Perhaps what I mean is more clear if you substitute "subentity of a way" > instead of "lane". Basically, if you have a way which consists of two > lanes, and those lanes are identical in every manner except that one is for > cars and trucks and the other is for cars only, I'm okay with somehow > labeling individual lanes within a single way. On the other hand, if one > lane connects with some roads, and the other lane connects with other roads, > and which is which is important, this needs to be represented as two ways, > not one.
Ah. So I think the issue here is whether a lane can be: 1) explicitly traced out as a "way" (i.e. series of nodes), or 2) assumed to follow the same path as the "parent" way If a lane is related to its adjacent lanes and "parent" way through the use of a relation, I think 1) is necessary. Right? Because the members of http://wiki.openstreetmap.org/wiki/Proposed_features/lane_and_lane_group, for example, are themselves ways (i.e. series of nodes)? _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
