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

[Ulf Lamping]

Dave Stubbs schrieb:
> I think the problem is the use of the language in regard to the
> feature itself. A single person approving of a tag is obviously fine,
> but once a vote happens, and about 10 people approve of it, the tag
> then becomes an Approved Tag... or at least some people think it
> does.. this obfuscates the rather limited source of approval.
>   
If there are 10 people thinking about a tag and no one disagrees, this - 
to my experience - is a *very good* indication that the tag is in fact 
well worked out. Usually it's the other way round, there are 10 people 
with at least 2 or 3 different points of view and even after the 
proposal was discussed for some time and it comes to voting more than 
50% of the votes just fail. So the approval is not that weak as you may 
seem to think.

And what's the alternative? Having no approval (or call it 
recommendation, or whatever, I don't care) at all and let the current 
mood of one of the developers be the best way to decide things?

To my experience a proposal gone through discussion *and* voting is 
usually very clear how to use and therefore needs much less discussion 
afterwards. And voting is one important reason of this output, as this 
points out if there are still disputes or open points once the 
"discussion dust" settled a bit.
> Deprecation has a similar but more annoying problem... a bunch of
> people on the wiki decide they don't use a tag or have a better way,
> so essentially disapprove of it. It then gets marked as "deprecated"
> to the complete confusion of the active mappers who are happily using
> it and actually approve of it but weren't around for the vote.
>   
Well:
1. deprecation really *rarely* takes place
2. I've only seen deprecation of tags that had real issues, so there was 
a good reason to change it

Yes I know, in six month from now the evil guys from OS will come and 
deprecate all of our current OSM tags in one big rush - and the whole 
project will fall into a big black hole ;-)
> I don't really think it's the language that's the main problem so much
> as the opaqueness and finality presumed in the end result
Sorry, but I can't see any opaqueness and finality. For each proposal 
you can read the discussion and voting - where's the opaqueness? If we 
find out in three months or so that a tag was a bad idea - we can easily 
start a new proposal and change it again - where's the finality?

There's a lot of criticism about voting. But unfortunately - and let me 
state that again - *no one* came up with a better way to find a good 
agreement that may last some time.

Regards, ULFL