[OSM-talk] Proliferation of path vs. footway

[Richard Mann]

I think the underlying problem with "path" is that it creates overlapping
definitions. Among data users there is a strong preference for tag
combinations to be hierarchical, and I think that preference is reasonable.
While having to deal with "doctor" and "doctors" is only a mild pain, trying
to deal with multiple overlapping fuzzy definitions for commonly-used tags
is enough to make your head spin.

So - path should either pitch itself to cover everything (ie footway should
be a subset of path), or to cover a clear niche (ie path should be
independent or a subset of footway). The deprecation of footway/cycleway was
voted on (by not many people, but nevertheless), and the deprecation was
rejected, but some people don't seem to be able to take no for an answer.

You can use the same analysis for footway/cycleway. Either one is a subset
of the other, or they should be clearly independent. The wiki tries to make
them independent ("mainly or exclusively"), but they aren't in many
countries, hence the confusion. I think treating "cycleway" as a subset of
footway is a more robust model, allowing the grey area between the two to be
described more accurately, rather than trying to pretend it doesn't exist.

Richard
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20090813/4b89ac2c/attachment.html>