[Tagging] Stop the large feature madness

Kevin Kenny kevin.b.kenny at gmail.com
Thu Apr 18 20:10:06 UTC 2019


On Thu, Apr 18, 2019 at 2:30 PM Christoph Hormann <osm at imagico.de> wrote:
> No, the concept of verifiability defines a clear path for resolving
> disagreement - you verify the information on the ground and if there is
> still disagreement it is by definition something that is not verifiable
> (because several mappers evaluating the situation independently do not
> consistently come to the same results).

In the specific case of names, we've invented name:language and
alt_name and old_name and name_1, name_2, etc. to deal with the cases
of, 'everyone agrees that there's a pond/mountain/building/whatever
here, but not all the locals call it by the same name.' The name may
be verifiable in that if you have a sufficiently large sample of
locals, you'll hear it, or if you ask, you may get the answer, "yes,
that's what some people call it."

I do recognize that you tend toward the 'strict verifiability' camp,
and that I've somewhat caricatured it by saying 'if a stranger dropped
into a location can verify everything about it by direct observation
without consulting the locals or outside sources.' That strict a
definition excludes a good many names, at least in the rural US, which
isn't big on hanging a sign on every named feature.

The idea that every object, even among those that can be easily mapped
in a day, has a single True Name, is simply an incorrect assumption
around here.



More information about the Tagging mailing list