[Tagging] Tagging type of ownership of a road

Jens Glad Balchen balchen at saint-etienne.no
Sun Apr 16 18:45:28 UTC 2023


On 16.04.2023 16:55, Greg Troxel wrote:

>> If we look at this from a data perspective, the most important
>> information for us to capture /today/ is which public entity type owns
>> the road and put this in the ownership tag. The specific entity can be
>> derived geographically with probably 100% accuracy. If we have the
>> specific entity available in a data set, we can put this in the owner
>> tag. If the operator at some point in the future again diverges from
>> the owner (like with the county roads), we can put that in operator.
> You need to define schema for the code points in "ownership".  They are
> far from obvious.

It seems this is already defined. The suggestion to use ownership=* 
wasn't mine, but looking at the wiki page it seems to fit quite well. Do 
you think the existing schema for ownership=* is missing something?

> You are defining a rule that one can use "ownership" and admin
> boundaries to find a value for "owner".  That's ok, but it needs to be
> clearly documented.  It's easy for a human to stick something in and
> think they have communicated, but it's much harder for a data consumer
> to get it right.
> You are defining a rule that "operator" should, if missing, be presumed
> to be the same entity as "owner".  That's ok, but it needs to be very
> clear.

Isn't it a general assumption for everything that the owner is the 
operator by default? I'm not saying "isn't it always true", but isn't 
that generally the assumption if no other information is presented?

Are you suggesting we expand the existing wiki page on ownership=* with 
that clarification (as it relates to roads, anyway)?

> Basically I am asking you to think through the situation more broadly,
> for realities not matching yours and data consumers other than the ones
> you are contemplating, and I feel this is too much of a hack for a point
> problem.  Maybe you meant all of what I am asking, but it seems very
> much too implicit.

What is a point problem?

Cheers,

Jens



More information about the Tagging mailing list