[Tagging] key damage and HOT

Andrew Harvey andrew.harvey4 at gmail.com
Thu Feb 6 13:38:31 UTC 2020


ruins:building=yes is not just tagging for other mappers, it's
accurately describing the feature on the ground, a ruined building. It's
not quite a building=yes, but not really nothing left on the ground, so
it's just part of the lifecycle.

On Thu, 6 Feb 2020 at 19:23, Paul Allen <pla16021 at gmail.com> wrote:

> On Thu, 6 Feb 2020 at 05:06, Warin <61sundowner at gmail.com> wrote:
>
>> It seems to me that the present use of the key damage=*, which has no
>> documentation, would better fit into the life cycle system.
>>
>
> If we need it to map damage all (I think we do) then we need both ways of
> doing
> it, for the same reasons discussed some weeks ago regarding disused=* and
> the disused: lifecycle.
>
> We need both because they have different effects.  They have different
> effects because we need those different effects.  I'll use "disused"
> rather than "damaged" below to make the point clearer.
>
> Lifecycle prefixes prevent rendering of the feature.  They are equivalent
> to
> deleting the feature tag and adding a note to the effect that the object is
> a disused <whatever>.  Except that the word "disused" might not appear
> in the note and a synonym or circumlocution might be used instead.  Having
> disused:amenity=hospital allows database queries to pick out hospitals
> (used or disused), disused hospitals, or functioning hospitals.  Removing
> the amenity=hospital tag completely prevents the object appearing in
> queries for hospitals, or for disused hospitals.
>
> Having disused=* doesn't prevent rendering.  A disused water tower looks
> like a functional water tower and is (usually) a landmark used for
> navigation.  Again, database queries can pick out water towers, disused
> water towers and functioning water towers.
>
> Is having two ways of doing it tagging for the renderer?  No more than
> having
> amenity=hospital or removing amenity=hospital.  One renders the object
> as a hospital and the other does not: the mapper chooses based upon what
> the object is.  Objecting to a mapper being able to decide whether it
> is rendered as a hospital or not means objecting to being able to tag
> a POI in any meaningful way.
>
> Isn't it recording history and OSM doesn't do that?  It serves two
> purposes:
>
> 1) QA.  A formalized way of telling other mappers that no matter what the
> POI looks like in aerial imagery, street-level imagery or a drive-by, the
> object
> isn't what it appears to be.  A note could do that, but is opaque to
> database
> queries ("former hospital," "was a hospital," "no longer a hospital," etc.)
>
> 2) A formalized way of telling data consumers who query the POI that it
> isn't what it appears to be.  Don't hang around that church you spotted and
> wait for it to open up so you can have a look around, it's disused.
>
> Will all renderers honour those interpretations?  Probably most will.  It's
> easy to not render tags with lifecycle prefixes by simply ignoring them
> as being unknown.  It's easy to render tags with disused=* by ignoring
> "unknown" tags.  A renderer would need extra code and have to be
> somewhat perverse (IMO) to render tags with lifecycle prefixes or
> not render POIs with disused=*.  We can probably rely upon these
> behaviours for most renderers.
>
> There are two ways of doing it, and ithat's a good thing.
> --
> Paul
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20200207/11d7e3fc/attachment-0001.htm>


More information about the Tagging mailing list