[Tagging] Re-using the marker tag for survey points

François Lacombe fl.infosreseaux at gmail.com
Wed Jun 16 21:37:25 UTC 2021

Hi all

Le mer. 16 juin 2021 à 01:40, Paul Allen <pla16021 at gmail.com> a écrit :

> On Tue, 15 Jun 2021 at 23:36, Kyle Hensel <K.y.l.e at outlook.co.nz> wrote:
>> Hi, a suggestion on the Survey Point proposal (
>> https://wiki.osm.org/Proposed_features/Survey_Markers) was to re-use the
>> tag marker=* instead of creating survey_point:structure=*
>> This sounds like a good idea, but there are some issues:
>> * Some values of survey_point:structure= are not valid for utility poles,
>> and vice-versa, so the marker=* tag would become complicated - some values
>> can only be used with certain tag combinations.
> Not good.  Some mappers will mistakenly use wrong combinations.

We should be careful about validity and probability.
survey_point:structure=* and marker=* are both about shape and appearence
of survey points or markers, not purpose.
If some values are not valid (understood as unlikely) with
man_made=survey_point or utility=*, why mappers be confused when they look
on ground?
The shape is enough to determine the right value to use, why have we to
look to purpose? Did we confuse shapes and purpose?

> * It will be difficult to create presets in editors like iD because of the
>> foregoing point – fetching a list of common values from taginfo wouldn’t
>> tell you which values are valid for which features.
> Because it would require special-case code to figure out which values are
> valid (yeah, table-driven but extra work, especially when new values are
> added) then iD's authors are likely to refuse to implement a preset for it.

Josm warns me very often about poor choices of values depending on context
(in a simple mapcss file). Why iD and other tools can't do the same?

Furthermore, Data Items now allow to state some values incompatible with

> * there will be a tagging conflict for survey points on the same node as a
>> utility pole, since it’s not clear what marker=* refers to (like the
>> service=* tag on roads/railways)
> I don't know if this happens in reality, but it probably could happen.

A survey point is precisely placed (and kept in place), an utility marker
or highway milestone is here to give an information (and can be moved a few
meters away without bother anyone)
Do you have any examples of survey point and utility marker found in the
same place?

> Another reason not to do this.
>> So what does this mailing list think of using the marker=* tag instead of
>> creating survey_point:structure=* ?
> I can only speak for me, but I think it's not a workable idea.

Do you feel the same about usage=*, location=*, tower:type=*, substation=*,
material=*, surface=* ...?
Will you agree to define a new key each time you need to specify a subset
of values in a given context?

Interesting point to discuss, thank you Kyle

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20210616/0a78bb62/attachment.htm>

More information about the Tagging mailing list