[OSM-talk] Classifying Ways worldwide

[Wollschaf]

On Thu, 03 Aug 2006 12:50:42 +0100, Etienne wrote:

> By all means have a width tag in metres, but only use it for cases where
> the width has actually been measured to at least +/-1m accuracy.  For
> other cases approximate_width=3m or subjective_width=narrow seems to me
> to be a better record what we are able to observe.

On Thu, 03 Aug 2006 12:03:32 +0200, Joerg Ostertag wrote:
> I would restrict the field to meters. Since this way we have a chance of
> making use of it in automatic processing. And give the user some hints
> to enter data...

Measuring width in vehicle dimensions is easier to remember while mapping.
At least I would be able to tag a lot of roads right now without having
made notes. Deriving a width value and entering that would make me feel
bad because I'm uncertain that it's really right.

It's hard to get the width of a four-lane motorway with breakdown lane
even approximately right... A width attribute should be included, and be
restricted to measured values. A guessed width is also good... but that
does not completely dismiss a (well thought out [1]) way classification
theme that works with vehicle dimensions.

For automated data processing, each road type can be assigned an
approximate value (as Joerg proposed), that is just derived by the
computer. Why force users to enter hard to comprehend numbers, while the
computer can just do the same?

If the road is categorized / classified, an approximate width value is
known for usage cases where this is needed. If an approximate width is
given, that is used instead of the derived value. The same for an
exact with value - which overrides the approximate width. 

Route planners can use the road classification scheme without having to
know road width. The classification from A to F (or whatever) includes a
little more information than just road width.

Wollschaf