[OSM-talk] stop deleting abandoned railroads

Lester Caine lester at lsces.co.uk
Fri Aug 14 23:03:45 UTC 2015

On 14/08/15 23:14, Mateusz Konieczny wrote:
> On Fri, 14 Aug 2015 22:51:13 +0100
> Lester Caine <lester at lsces.co.uk> wrote:
>> But the
>> one thing that the current model has got the capability of handling is
>> start and stop dates for any facet of an object from the name of a
>> shop to the evolution of the road and rail system over time. That the
>> 'main' database only displays elements which have not yet acquired a
>> stop date is how the model currently works, while the OHM version
>> simply maintains multiple time stamped versions of the same data.
> This is not true. Current model - from mappers, through programs used
> to edit to data consumers is unsuitable for handling several
> intersecting objects, verifiable only for specialists.

? That is basically what I said ... but the problem of combining macro
and micro mapping has been discussed without much progress.
Adding a stop_date is the only element that works ... and is the correct
action rather than 'delete'.

>> Rather than having to recover the data from the change log ...
> Date of deletion is at best loosely correlated with date of
> destruction of represented object. In particular most edits either are
> mapping the same objects in greater detail - despite lack of change on
> the ground, adding new objects present in reality since decades or in
> some cases thousands of years (in case of natural forms it may be even
> higher).

Start and stop dates are verifiable facts that can be recorded ... WHEN
that information is added is only of academic interest. Potentially the
movement of the techtonic plates could be mapped , and certainly recent
shifts due to earthquakes have a date when the variation occurred, even
if it then takes many months to actually map the movement.

> Many of deleted objects never represented reality.

No argument on that. What I am talking about is exactly where the
historic material has been mapped in detail, but new road works result
in a change the is subsequently also mapped. Or new roads are added with
start_date tags which allows the evolution to be recorded. Reality
changes over time?

> Mapping all buildings that ever existed may seem simple for some cases.
> But many cities were destroyed (partially or fully) and later rebuild.
> I would delete any encountered OSM elements that are marking objects
> that are fully and completely gone (obviously, I would do it only for
> places where I can verify this and after asking original mapper for
> clarification whatever there are some mappable traces - maybe just
> tagging was wrong). One exception would be for elements representing
> recently destroyed objects that were mapped to protect from recreation
> based on outdated aerial images (in that case I would convert it to
> ways marked with note=* asking to avoid recreation without survey).

Tag when the object ceased to exist in that format with a stop_date. The
action date is of little use for the historic record. Visibility of the
object is a different problem.

Lester Caine - G8HFL
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

More information about the talk mailing list