[Tagging] improve tagging of traffic_calming

[Gerd Petermann]

Chris Hill-6 wrote
> On 20/11/15 16:47, Gerd Petermann wrote:
>>
>> So you think that every node tagged traffic_calming=* should
>>
>> additionally be tagged with highway=traffic_calming?
>>
>> That would mean  we have >99% traffic_calming nodes with
>>
>> incorrect tagging.
>>
>>
>>
> I think you misunderstand OSM. Your assumption of incorrect tagging is 
> wrong. There is a diversity of tagging. That is how OSM works. Some will 
> say 'how will data consumers deal with this muddle?'. That question is 
> usually asked by people who have not tried to use OSM data. People who 
> use OSM data know that it *always* needs processing to be useful for the 
> particular use it is being put to. Dealing with that diversity is part 
> of this processing and not especially difficult nor does it add much to 
> the process. Anyone who has written such a process to do this then has a 
> template to do this again easily.

Sure, as a co-developer of mkgmap (1) I am very familiar with the process
of consuming OSM data. I also understand that it requires years to establish
a
new tagging scheme. 
I think one problem here is that you list these arguments for a change that 
happened far back in history. 
1) Most of the ~100 nodes with the 
highway=traffic_calming and no traffic_calming=* tag were last changed 
before 2014, 
2) most of those nodes with both highway=traffic_calming and
traffic_calming=*
were first mapped with highway=traffic_calming and later the tag
traffic_calming=*
was added.
3) nearly all traffic_calming=* mapped in 2015 do not have the
highway=traffic_calming=*
tag.
I don't want to stop or change this process (although this discussion seems
to lead that way).


Chris Hill-6 wrote
> Diversity in tagging is not a problem. Mechanical edits that destroy 
> detail or upset precious mappers who have their carefully chosen tagging 
> replaced by a remote person are a problem.

Well, of course diversity in tagging is a problem. For each data consumer
and
for each new mapper who tries to find out how to map something so that their
preferred OSM-data-consuming-program is understanding the data.
The problem exists and causes lots of work. 
Diversity in tagging is also a basic feature of OSM and I think it is a good
way to
find out how real world elements can be described with abstract attributes.

My understanding is that this is a process and that the result of the
process 
can be a well established tagging scheme with few special cases  - I think 
traffic_calming=* is one of those - or total chaos as we have with 
highway=path/footway/track  or residential/unclassified.

(1) http://www.mkgmap.org.uk/



--
View this message in context: http://gis.19327.n5.nabble.com/improve-tagging-of-traffic-calming-tp5860596p5860668.html
Sent from the Tagging mailing list archive at Nabble.com.