[Tagging] R: Relationship between place=* and name=*

Michael Montani michael.montani at un.org
Thu Jan 28 14:16:18 UTC 2021

I agree with such a resolution.

To me the tag place=* is most used to target isolated dwellings, hamlets or villages as such, which most of the times may have a name but many times, in some geographical contexts, may be lacking of it or may be inherited by temporary socio-economic factors like patronimics and jobs (which would require a missing name rather than a noname=yes).
Anyway from a semantic point of view, to me, the settlement information is somewhat unrelated to its toponimy, expressed by name=*, and almost always present but not mandatory.

Most of the places we mapped in Africa all these years doesn't have a name, and probably will be difficult to ever get this information. Many times small settlements are appearing, moving or disappearing from the map for many reasons (attacks, nomadic populations, climate change..) such that we are usually mapping them as nodes and residential areas, rather than mapping every single building / hut. And that's as Rafael was saying, an established practice.


Da: Rafael Avila Coya <ravilacoya at gmail.com>
Inviato: mercoledì 27 gennaio 2021 10:20
A: tagging at openstreetmap.org <tagging at openstreetmap.org>
Oggetto: Re: [Tagging] Relationship between place=* and name=*

Hi all,

If I understand it well, the matter that originates this thread is that
the place=* tag wiki (not the Places wiki by the way) is confusing in
its introductory sentences, as it may be interpreted by many mappers
that a name=* tag has to always be added to the place node (or a
noname=yes in case we know that the place doesn't have a name, that is
quite different from not knowing the place name or its lack of it).

Thousands of hamlets and villages and even towns have been mapped with a
node and without a name all around the world, using armchair mapping,
both by locals and non-locals. This is a common practice since almost
the beginning of the osm project.

Mapping places as nodes without the name=* (or noname=yes) tag gives
already an information. It visually helps to check the connecting roads
between those settlements that are missing, and its classification.

Place nodes are independent of landuse of the area (mainly residential),
and are something easier to map than landuse, so I may add place nodes
to unmapped settlements and leave the (for example) residential areas,
building footprints and other info for later on, or for other users to
do it. We do this all the time, everywhere.

I think we should focus on how to improve the wording in the place=* tag
wiki to avoid that misunderstanding. Any suggestion?

For example: we could add the tense

" Adding the actual name=* of the place is not compulsory, although
highly recommended when you known it."

to the tense

"Used to indicate that a particular location is known by a particular
name, to indicate what sort of "place" it is. A place tag should exist
for every significant human settlements (city, town, suburb, etc.) and
also for notable unpopulated, named places.".

Or something similar.



O 26/01/21 √°s 18:48, Michael Montani escribiu:
> Hi all,
> I wanted to ask one doubt on the usage of place=* nodes with name=*,
> especially when mapping villages in Africa. Is it mandatory or highly
> suggested?
> The wiki definition of place=*
> <https://wiki.openstreetmap.org/wiki/Key:place> states:
> "Used to indicate that a particular location is known by a particular
> name, to indicate what sort of "place" it is. A place tag should exist
> for every significant human settlements (city, town, suburb, etc.) and
> also for notable unpopulated, named places."
> which in my opinion is like saying human settlements cannot exist if
> they don't have a name. The general practice I often saw when mapping
> human settlements in Africa is to add place nodes within residential
> areas (with exception of (sometimes) place=isolated_dwelling and
> (always) place=locality), whether or not the name=* is known.
> Indeed when mapping rural settlements, especially from satellite images,
> toponimy information cannot be retrieved even though human settlements
> are there. Furthermore, wrt to landuse=residential areas, place=*
> carries approximate population information (which is quite useful and
> well established in humanitarian mapping).
> Many times rural settlements in Africa have names:
>   * which are coming from family chiefs (patronimics) or important
>     people that lived there (ex1
>     <https://www.openstreetmap.org/node/3097654804>, ex2
>     <https://www.openstreetmap.org/node/3097654826>) (same as some
>     places in EU / US though!)
>   * which describe the purpose of the settlement or job of the people
>     living there (ex3 <https://www.openstreetmap.org/node/7503479607>,
>     which translated from Hausa would be 'Yan Gero Along the House of
>     Saba, the Seller of Dawa', with 'Dawa' being a local food)
>   * sometimes missing, but can be inferred from the names of amenities
>     (eg schools) which are coming from imports (like ex4
>     <https://www.openstreetmap.org/node/7777758115>)
> and, as you can imagine, many times they couldn't have a name at all,
> especially place=isolated_dwelling, as they would be described in other
> ways (like people/clan living there or landscape features to navigate up
> to that place).
> My question is not a matter of rendering, but about the definition of
> place in the wiki, and whether you believe or not name=* could be
> sometimes skipped when mapping places (try to run
> 'node["place"]["name"!~".*"]({{bbox}});' on Overpass anywhere in
> Africa...) or the wiki definition should be revised.
> Thanks,
> Michael
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

Tagging mailing list
Tagging at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210128/48f74fa2/attachment-0001.htm>

More information about the Tagging mailing list