[Tilesathome] Large Maplint files in USA

[Knut Arne Bjørndal]

On Tue, Jan 06, 2009 at 02:44:42PM -0500, Matthias Julius wrote:
> Patrick Kilian <osm at petschge.de> writes:
> 
> > Hi,
> >
> >> There's a bulk import happening of data in Massachusetts, USA, and
> >> Maplint is flagging every single node with not-in-map-features because
> >> of the "attribution" tag applied to these nodes (see
> >> http://server.tah.openstreetmap.org/Browse/?layer=maplint&z=12&x=1239&y=1518
> >> for an example.) This causes Inkscape to use enormous amounts of
> >> memory when rendering these tiles -- last night, I had a render fail
> >> because the Inkscape processes rendering z12 and z13 managed to fill
> >> all 6 GB of RAM+swap on my system.
> > If the "attribution" tag is common enough to cause that problem it
> > should just be listed and explained in map features. If you put it in
> > feel free to poke me. I'll run the rebuild the maplints "is in map
> > features" list and commit it to svn.
> 
> If a mass import introduces a new tag I would not call it common even
> it is in the database a million times.
> 
> Since OSM promotes free form tagging not-in-map-features is bound to
> produce a lot of false positives.  It might be a good idea to have an
> ignore list besides Map Features on the wiki.

There is already a ignore list, in the source of the perl
script. Since it requires SVN access to modify the test anyway I don't
see why we don't simply add one more tag that way instead of doing yet
more ugly wiki-parsing.

I think there's a bit of a difference between things like tiger:* and
attribution though, it sounds like a reasonable tag to me...

(but NO! don't start another voting thread)

-- 
Knut Arne Bjørndal
aka Bob Kåre
bob+osm at cakebox.net
bobkare at irc
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 4167 bytes
Desc: not available
URL: <http://lists.openstreetmap.org/pipermail/tilesathome/attachments/20090106/fd298050/attachment.bin>