[Tagging] waterway=lock_gate - is it only for nodes?

Yohan Boniface yohanboniface at free.fr
Tue Mar 17 17:16:01 UTC 2015

On 17/03/2015 16:16, Malcolm Herring wrote:
> On 17/03/2015 13:19, Martin Koppenhoefer wrote:
>> their tagging scheme is somehow a parallel universe,
> Not parallel, but in some cases complimentary. For 95% of objects there
> are no overlaps with standard OSM objects. It is those 5% of cases that
> do overlap that people notice & hence assume that OpenSeaMap have an
> alternative set of tags for everything.

Those 5% are still an issue.
Redundancy in the data, even if not wanted in theory, is part of the 
life of OSM because of many reasons, that we can resume by saying that 
the community is diverse.
Though, in the case of OpenSeaMap, the fact that *all* keys that overlap 
are duplicated shows, in my opinion, that there is no intention to 
*integrate* with the community. I understand that having one coherent 
namespaced scheme for all the keys interesting OpenSeaMap makes life 
easier for data processing, but, imho, that's more of parasite than 
nutrient for OSM data.
Plus, remember that, from time to time, we spot some mass edit changes 
where OpenSeaMap people **replace** OSM community tag with it's own 
namespaced version.
Each data consumer needs to adapt to the database, not the database to 
each data consumer.

So yeah, 95% of OpenSeaMap objects are specific and not overlaping with 
standard OSM objects, and this in fact is actually *just* OSM data 
related to sea and boat navigation, and this is *nice*. But still, after 
years of debate, there is friction on the other 5%, and that's sad, 
specifically because this devalues the OpenSeaMap perception in the 
community, and thus the OpenSeaMap data.

As long as you will try to control the data (and scheme) from a central 
point, you will loose the power of the community.


More information about the Tagging mailing list