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

[Gervase Markham]

Frederik Ramm wrote:
 > Even without a mechanism of "approval" people will eventually settle
 > down on a small number of ways of tagging common features. Not having an
 > approval system does not mean people stop talking.

No, but it does mean that people don't re-re-visit the same decision 
again and again, because there is a formal way to say "the decision is 
made" and a common place to record the rationale.

 > To give an example, I was recently looking for a tag for "vineyard".
 > Browsed the wiki and found a longish discussion about the advantages of
 > "landuse=farm,produce=grapes" over "landuse=vineyard" or vice versa.
 > Settled for "landuse=vineyard" and even updated Osmarender to render it
 > properly.

Right. So the rendering rules (and any other data consumer) get at least 
twice as big, maybe more, as each renderer has to deal with the 2, or 3, 
or 4 common ways of tagging a particular feature. How is this an 
improvement?

 > So as long as there's at least some discussion that can be found (either
 > in the Wiki or on the lists), things will go nicely even without formal
 > proposals and approval. A plain simple "hey folks, how do you tag <x>"
 > plus the 10 mails that follow is enough. Either all 10 use the same
 > anyway (or 9 of 10 do), then the issue is clear, or if every single one
 > of the 10 uses something else, but an approval/voting process wouldn't
 > lead far then.

Well yes, it would, if the ten people had a discussion, looked at the 
pros and cons of each method and decided which was the best and most 
general.

Of course, if everyone dogmatically says "I tag this way, and nothing 
will ever change me" then an approval/voting process doesn't help much. 
But I don't think there are many OSM members who are that unreasonable 
and dogmatic.

Gerv