[Tagging] Solving the tag chaos (was: Re: Government buildings)
stevagewp at gmail.com
Mon Dec 7 13:07:47 GMT 2009
On Mon, Dec 7, 2009 at 11:58 PM, Mike N. <niceman at att.net> wrote:
> The tagging reference procedure is definitely grown beyond easy usage.
> An interactive tagging reference resource like the John Smith's demo site
> from a short time ago is on the top of my "most wanted list".
What's the URL for that?
I care most about simply having a centralised resource that everything else
trickles from. At the moment it seems that the four sources of tag
information (wiki, renderers, editors, actual usage) can all feed off each
other in any arbitrary direction. It's quite bizarre really. Almost any
combination is possible: Potlatch implements a tag because JOSM does. One
mapper uses a tag because he saw it on a wiki. Someone documents a tag on
the wiki because he saw it in Tagwatch. A mapper uses a value because he saw
someone else do it.
A few suggestions:
1) Document in the 'map features' list whether each of the main editors and
renderers supports it.
2) Provide a standard list of tags and values with standard definitions,
which all editors should use for their autocompletes/dropdowns.
3) Allow renderers and editors to support "unofficial" tags, but encourage
them to document this fact on the wiki, then quickly launch a process for
confirming, modifying, or rejecting the tag.
4) Periodically perform mass reconciliations of tags.
It's fine that Mapnik and Osmarender (and others) render the same data
differently. Having them *interpret* the same tags differently doesn't seem
to have any benefits though. Having lots of people randomly use various tags
not recognised by any renderer (or router) seems like a complete waste of
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Tagging