[OSM-talk] Wiki: chriscf vandalism

Richard Fairhurst richard at systemed.net
Sat Jan 31 11:47:57 GMT 2009


(Nop's e-mail went to me rather than the list but I'm guessing that  
was a mistake - and he probably expressed the other side best)

Nop wrote:

> I would consider it the basic principle of democracy/a community  
> that things established by vote need to be changed by vote, even if  
> the need for change is obvious.
>
> I do not agree with the tag either, but as I sort of believe in  
> democracy I strongly oppose the overriding of votes by individuals.

Well, this is the crux of it. I'm not convinced the form of democracy  
we have in the tag voting is at all helpful.

The problem is that people vote on tags:

- without knowing anything about the subject
- without ever having mapped the feature in question
- without any intention of ever mapping the feature in question

which makes the votes meaningless. How does it help for me to cast my  
vote on (say) amenity=baby_hatch? I've never encountered one and I  
doubt I ever will. But there are probably experts on baby hatches  
within the OSM community who can make an informed decision on how it  
should be tagged. Why should me and my mates be able to veto that?

The voting did once mean that, even if the tags didn't benefit from  
"subject knowledge", they did benefit from "OSM knowledge". In other  
words, even though the voted tags might not correlate much to the  
real world, they were at least reasonably consistent within an OSM  
framework.

With smoothness that's gone out of the window. As far as I'm  
concerned, with the approval of smoothness=very_horrible (come  
_on_!), all bets are off. The voting system has just voted itself  
into irrelevance.

> Following your thoughts, as this is my conviction, I should stand  
> up to it and immediately undo Chris' "illegal" changes, thus  
> starting an edit war?

No need, there's been an edit war for about two months now. ;)

> I would rather suggest tackling the real problem with the voting  
> system or at least re-open discussion and vote of a badly designed  
> tag.

Oh, absolutely, I agree. We should. In the meantime Chris is the only  
person actually doing something about it while the likes of me just  
mither about how things should be better.

But, given that this is a good opportunity to start thinking about it  
seriously:

I'm starting to wonder about a "Tags I Use" system. In other words,  
if I think I have a smart way of tagging tracks (their surface, their  
cyclability, conditions through the year, etc.), I document it -  
maybe on the wiki (/User:Richard/Tags_I_Use), maybe someplace else. I  
explain what I use, why. Other people do the same.

A miraculous aggregator then goes through all these pages, drawing in  
some Tagwatch data, and reports "50 people are using surface=gravel,  
10 people are using smoothness=very_horrible, 1 person is using  
my_bike_suspension=knackered" - and links to people's documentation.

Then, for those who like to have everything in a central place, once  
the tags have been used n times, they can go in Map Features.

That would be a really, really useful tagging resource - one based on  
real-world usage and knowledge, not on a very small number of largely  
uninformed votes.

cheers
Richard




More information about the talk mailing list