[Tagging] bicycle=no

Roy Wallace waldo000000 at gmail.com
Tue Dec 8 03:25:22 GMT 2009

On Tue, Dec 8, 2009 at 11:21 AM, Steve Bennett <stevagewp at gmail.com> wrote:
> On Tue, Dec 8, 2009 at 11:41 AM, Roy Wallace <waldo000000 at gmail.com> wrote:
>> I strongly disagree. The surveyor must only tag information that is
>> verifiable. They should NOT tag
>> "I_think_the_router_should_send_bikes_down_here=yes". Notice the
>> "I_think" - that means it isn't verifiable.
> No, but
> "100_bikes_come_through_here_every_saturday_morning_and_arrests_so_far=0"
> is.

I'm not sure what you mean - do you understand my point that
subjective assessment of path suitability for a "bicycle" is not a
good way to tag, because it is not verifiable?

If you do understand this point, and if you still want to "map on the
basis of common practice", then do you have any ideas on how to do
this, *with verifiable tags*? I'd be very interested! (I'm assuming
your use of the key above was sarcastic :P)

>> In other words, the person "best placed" to set the criteria for what
>> they want to ride on is *the user*.
> The user don't got no tools to let him make use of all this surface
> information.

Don't tag on the basis of existing tools. It's easy to imagine a
routing application that would allow a cyclist to select whether they
want to go off-road (surface=unpaved, dirt, etc) or want to commute to
work (surface=paved, etc.). If there's a market for it, and if the
data is in the OSM database, routing applications will follow (same
analogy for renderers...).

> Mapping millimetre by millimetre the surface of a city is less
> useful than, say, mapping places where bikes can go.

Even if it is more useful, "places where bikes can go" is not
well-defined. This is the whole point of this thread. If you can come
up with a more specific definition (e.g. based on law, or physical
characteristics, or whatever it is that you mean), and a set of
verifiable tags that are able to express this, then you'd have my

On the other hand, even if you argue that it is *less* useful,
surface/width *is* well-defined, and verifiable, and able to be used
for routing, renderer, etc...

More information about the Tagging mailing list