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

[Dave Stubbs]

On Wed, Mar 19, 2008 at 2:03 PM, Gervase Markham <gerv-gmane at gerv.net> wrote:
> Frederik Ramm wrote:
>  > Why not ditch the whole notion of "approved" features altogether. It
>  > doesn't cut any meat in our community anyway. What does "approved"
>  > mean, and who has the right to "approve" something?
>
>  Having an approved set of tags means that there is ideally 1, but
>  certainly a small number of ways of tagging common features, rather than
>  15 or 50. This makes it much easier for renderers, routing and other
>  types of software, and much easier for people who are improving an area
>  of map that someone else has worked on to figure out what they meant.


Just because you don't "approve" the tag, doesn't mean everybody does
their own thing and ignores everyone else without any discussion.
And this thread is starting to give me deja vu.... so lets not rehash
this argument again.


>
>  It also means that when a particular tag is used, it only has one
>  meaning. Without some standardisation, does maxspeed=50 mean mph or kph?
>  Or does it vary from country to country?

whoa! now that's some serious deja vu....

>
>  What is the difference between your argument and "Why have the notion of
>  an "approved" set of HTML tags? The web is a collaborative community.
>  No-one has the right to approve anything. We should all just use the
>  markup tags that seem most sensible."?

it's called XML... and microsoft have proved what really cuts the mustard there.

>
>
>  > Right, generate it from the planet file and that's that. Maybe have a
>  > wiki page that documents what the renderers do and at what zoom level
>  > (ideally auto-generated as well).
>
>  Except that such a generated page would have no way of ordering and
>  classifying the tags so that you could find the one you wanted.

Here I agree. You want some more information than just that. ie: a
system where people can document and add meta data to tagging schemes
such as categorisation and common groupings (although maybe that one
can be automated a bit). A planet dump is a good base, but some
explanation for horse=yes wouldn't go amiss.

Dave