[OSM-talk] Tag proposal/approval system is too heavyweight

Richard Fairhurst richard at systemeD.net
Thu Mar 20 09:11:08 GMT 2008


Ulf Lamping wrote:

> Frederik Ramm schrieb:
>> Well. Before there were no vineyards on the map. And the discussion
>> was rather dead. What do you expect me to do, start a vote or just do
>> it in a way that works and gets vineyards on the map
> Well, I'd expect you to start a vote just like anyone else and NOT
> "abuse your power position" as one of the persons who knows how to
> change the rendering rules.

Well, there's an issue.

At present there are perhaps three main influences on what tags are  
used: what's documented, what's rendered, and what the editors  
present as presets.

I can only speak for part of the last-named (and arguably least  
significant). Potlatch's preset and autocomplete tags are meant to  
reflect what the community wants and uses. At present they are taken  
almost directly from Map Features.

Unfortunately Map Features may be diverging from the community. Tags  
are proposed and voted on by a very small subset of people. Commonly  
used tags have had their descriptions "clarified" so that they mean  
something significantly different from what they did originally. And  
we have the situation where someone (like Gerv) proposes a sensible  
set of tags for an area where he clearly has some subject knowledge -  
inland waterways - and where his proposals are being niggled and  
criticised unjustly by people who don't; which hardly encourages  
people with subject knowledge to lend their expertise. All the  
bureaucracy rather reminds me of Wikipedia, to be honest.

Consequently I am leaning towards replacing the Map Features-derived  
presets in Potlatch with something produced by a tagwatch-style  
approach (though retaining Map Features data as an option for those  
who wish to use it).

Is this abusing a position? No, I don't think so. It's the voting  
that has diverged from the community, not the other way round, and  
Potlatch should follow the community.

cheers
Richard




More information about the talk mailing list