[OSM-talk] On the ground rule on the wiki

[Anthony]

On Mon, May 31, 2010 at 1:57 PM, Frederik Ramm <frederik at remote.org> wrote:

> Hi,
>
>
> Anthony wrote:
>
>> By these definitions, something that is able to be confirmed as true or
>> false in an official online source is actually *more* verifiable than
>> something written on a street sign in a place where Google Street View has
>> not yet visited.  It certainly is verifiable, and it is not necessarily "on
>> the ground".
>>
>
> Something that is available from an official online source but not
> verifiable on the ground should not - in my personal opinion - be included
> in OSM.
>
> For the simple reason that we cannot improve the data - how should we if
> there is not reference on the ground? So the data will just sit there and be
> left to rot, or left to wait for another update by those who keep it. But
> OSM is not a "mirror" for official data. I don't want data that OSMers
> cannot work with; such data would only be in OSM for ease of retrieval, and
> I don't view OSM as some data dumpster for the world's geodata.
>

Well, I think there's a difference between being "verifiable on the ground"
and having "reference on the ground".  The name of a lake or a river has
"reference on the ground", even if the name of the lake or river isn't
printed on a sign.  A county border which is defined with reference to
roads, rivers, fences, etc. has a "reference on the ground", even if the
name of the county isn't printed on any of those features.

I can agree that OSM should not include data which has absolutely no
reference to physical (relatively) non-movable features on the ground.  But
once the feature is there, when is it okay to tag it with features which
aren't strictly "on the ground"?

Maybe the answer is "never".  Personally I think the correct answer is more
like "usually not", at least "usually not" in highly developed areas which
tend to put signs on everything.  But I think it's confusing if you refer to
that as encompassed by the "verifiability" rule.  It's really a separate
rule altogether.

On Mon, May 31, 2010 at 2:33 PM, John F. Eldredge <john at jfeldredge.com>wrote:

> Well, some people in the traffic-jam discussion seem to be taking the
> viewpoint that if something is not verifiable by people in other
> geographical locations, without actually visiting the location under
> discussion, then it should not be classified as being verifiable at all.
>

Huh?  So in any location where we don't have good aerials and street views
we can't map at all?  Or do photographs count (in which case, just take a
photograph of the traffic jam)?

Actually, wrt the traffic jams I'd just as well they not be in OSM, for much
the same argument that Frederik just presented ("the data will just sit
there and be left to rot, or left to wait for another update by those who
keep it").  But as far as I'm concerned, so long as people use tags and
features which don't collide with the things I'm interested in, I really
don't care what unmaintainable crap they add to the database.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20100531/fc8b21db/attachment.html>