[Tagging] Should 'highway=incline[_steep]' be discouraged?

Matthias Julius lists at julius-net.net
Wed Dec 30 15:43:27 GMT 2009


Roy Wallace <waldo000000 at gmail.com> writes:

> On Wed, Dec 30, 2009 at 3:01 AM, Matthias Julius <lists at julius-net.net> wrote:
>>
>> If you know the actual incline you can tag it with its value.  If you
>> have to estimate it anyway then a hard definition on what is steep is
>> not worth that much anymore.
>>
>> It is a subjective classification - not more and not less.  One could
>> think it implies a "FIXME: check for actual value".
>
> I have no problem with using a FIXME. But performing even a
> "subjective classification" still requires criteria as to how to make
> that classification. I.e. it is steep if ______. If you want to define
> steep by saying, on the wiki, "incline=*_steep means the mapper
> personally thought the incline was steep", go ahead - but IMHO that's
> not a good definition.

OK, that certainly makes sense.  It still is an estimate.

So, what is steep then?  15% or more?

>
>>> What I mean to say is, unless you can think of a better way to tag
>>> incline=* on nodes, we should encourage the use of incline=* for this
>>> purpose (as opposed to some other tag like inclination=* or
>>> steepness=*, or node_incline=* etc., which would just be silly).
>>
>> I would say all the incline tags should be moved to the ways.
>
> As Anthony said, this isn't equivalent, unless it's moved to a new way
> with infinitesimal length.

or to a relation that indicates that a way has an incline at a certain
node.  You could call this overkill, of course, but it has the advantage
to be more robust when the data is changed, especially when another way
is added to that node.

Matthias




More information about the Tagging mailing list