[Tagging] Tagging Voting system- time for reform?

johnw johnw at mac.com
Sat Jan 24 03:31:40 UTC 2015


> On Jan 24, 2015, at 11:04 AM, Richard Welty <rwelty at averillpark.net> wrote:
> 
> On 1/23/15 8:37 PM, Warin wrote:
>> 
>> Yes .. it makes the admin more complex. But it will get some to say something, and get others off the group. Flame away. 
>> 
> i do not think it appropriate for the membership of this group
> to set these sorts of parameters for controlling its membership. it
> goes against the grain of OSM as a project.

I think any body that dedicates itself to managing something aught to actually manage it - I think that is the reason this came up. 

But I think there is an even better reason that a lot of members don’t vote on proposals, besides a lack of enthusiasm in regards to what the tag fleshed out -

a Lack of domain knowledge coupled with a lack of a (somewhat) rigid tagging schema that can be applied across similar groups of tags. 

For example - the discussion raging over semicolon delineated tag values vs sub-key values is something I have no relevant experience in, and I couldn’t comment on it, let alone feel comfortable voting.

Similar to the water tap issue - I voted for the water tap because I want a way to tag taps - but the issues  coming up now about it breaking compatibility of the dataset is something I similarly don’t know, and I will refrain from commenting now.  Same with the Kiln questions in Tibet. 

But if there ware a more uniform tagging scheme then it would probably be obvious to a majority of the list members if a proposal *at least* properly fits into the format of tagging for a certain class of object  (bus routes, water taps, and buildings are all going to have different schema, of course) but there are several classes of tags where there is no set “standard” on how to implement the class, which makes proposals in that class a nightmare because it just devolves into what implementation schema should be followed. Since there a lot of older established tags and schemes that don’t follow more recent patterns, it just be comes a quagmire of what of all possible schemes something could fall under. 

My recent proposal of Landuse=civic is a good case in point - does everything get it’s specific landuse from amenity=* , like a hospital or a school? does it get’s a basic idea of purpose from the landuse area, like residential or commercial land? or is the idea of separating out governmental/civic amenities disliked - and the only big distinction should be between civilian and military? Handling police, fire, judicial, penal, and governmental building’s landuse becomes a fight over what scheme is better - or what key scheme is best, and no one can agree on that. - a courthouse isn’t commercial land - a police station isn’t a residence, and a City hall is more than Just a building. 

Those basic questions hamstring discussions - then coupled with how the changes will affect the dataset means people will inherently shy away from voting on proposals - or proposals will languish because there can be no definitive answer on something like landuse, since there are two basic kinds of tags, and I don’t think anyone wants to depreciate amenity=hospital from it’s landuse duties, nor landuse=industrial. 

And most taggers probably don’t understand the intricacies of supporting semicolon delineated values, nor kilns in tibet - so it makes it hard for everyone to vote. 

Opinions of the noob

Javbw






More information about the Tagging mailing list