[Tagging] Roman roads - was Re: "part:wikidata=*" tag proposal for multiple elements connected to the same wikidata id

Paul Allen pla16021 at gmail.com
Sun Sep 15 18:20:16 UTC 2019


On Sun, 15 Sep 2019 at 18:40, Martin Koppenhoefer <dieterdreist at gmail.com>
wrote:

>
> I see. Up to now, I only have mapped roads (or fragments) that were
> clearly original (or well arranged so that the layman would believe they
> were original), particularly with ancient paving (but maybe not in the
> original configuration, I have had my doubts sometimes, because of the
> traces of usage which weren't well aligned). If the same tag can also mean
> the approximate location of a former roman road of which there is zero
> visible trace, then it somehow blurs the mapped existent roman roads. I
> believe we should use 2 different tags.
>

>From what I've seen done by others, there are modern roads that follow the
old Roman road
which are tagged as historic=roman_road as well as the normal road tags;
there are tracks
which follow the original Roman road which are tagged as
historic=roman_road and as a track;
and there are places where the old Roman road used to be but nothing is now
visible which
are not mapped or tagged at all.  The new roads that follow the old Roman
roads, together with
the tracks that were Roman roads are then gathered together into a
relation.  I think you could
perhaps use the abandoned: prefix where there is solid evidence of where a
Roman road
was but nothing is there now, but that would annoy some people (don't tell
them about all
the abandoned railways that have been mapped).

>
> As I am mapping in Rome, by your criteria, I would probably have to add
> historic=roman_road to almost all roads in central areas ;-)
>

It's optional, not mandatory. :)  Also, don't forget that "All roads lead
to Rome."
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20190915/05777f1d/attachment.html>


More information about the Tagging mailing list