[OSM-talk] TIGER / Mapnik suggestion

[D Tucny]

On 05/09/07, Jon Burgess <jburgess777 at googlemail.com> wrote:
>
> On Tue, 2007-09-04 at 20:03 +0100, Robert (Jamie) Munro wrote:
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > Jon Burgess wrote:
> > > On Tue, 2007-09-04 at 13:34 +0100, Artem Pavlenko wrote:
> > >> On 04/09/07, Robert (Jamie) Munro <rjmunro at arjam.net> wrote:
> > >>> -----BEGIN PGP SIGNED MESSAGE-----
> > >>> Hash: SHA1
> > >>>
> > >>> Could the TIGER data (here http://sr71.net/~dave/osm/tiger/counties/)
> be
> > >>> short-cutted into the Mapnik DB by running all the OSM files through
> > >>> osm2pgsql?
> > >> Sure. I was thinking about loading some TIGER/Line (perhaps WA)
> before
> > >> FOSS4G. In fact we can have two PostgreSQL - one for planet dump and
> > >> other for TIGER, AND. It won't take years to setup and hopefully
> it'll
> > >> speed up loading data into main OSM db ;)
> > >>
> > >> Then people could see the data, and if they wanted to edit
> > >>> it, they could request their area gets put in the TIGER queue.
> > >> Yes, cool idea!
> > >
> > > Do we want to import and render these tiles on the same server as the
> > > existing mapnik layer?
> >
> > I'd vote yes.
> >
> > > The tiles could appear as 2 distinct layers (main OSM or Tiger).
> > > Alternatively I could merge the data during the rendering to give one
> > > layer with just OSM and another with OSM+Tiger.
> >
> > Ideally make one layer with the OSM data + Tiger data that hasn't been
> > imported already. I don't think we need an OSM without Tiger layer. T at H
> > will provide that.
>
> Mapnik rendering of about half the Tiger data is on my web server at:
> http://www.jburgess.uklinux.net/tiger-mapnik.html
>
> This is just the plain Tiger data I have not combined it with anything
> from the OSM planet dumps.
>
>         Jon


It's looking good!

I have noticed some oddness as a result of this...

Things like this...

http://www.jburgess.uklinux.net/tiger-mapnik.html?lat=5184672.34309&lon=-7920818.78055&zoom=14&layers=B

Shows a small chunk of motorway... I think, from the higher zooms, that it
might actually be a motorway link...

in google maps
http://maps.google.com/?ie=UTF8&ll=42.3506,-71.154521&spn=0.007422,0.021629&z=16&om=1

it doesn't... Here...
http://maps.google.com/?ie=UTF8&ll=42.349208,-71.153646&spn=0.000928,0.002704&t=k&z=19&om=1
on the sat view, it also doesn't look very motorway like...

There seems to be a good few of these scattered around Boston, is this some
mismapping of a type of link road to motorway link in the conversion?
Infact, looking at other examples, it almost seems that all link roads are
showing as motorway link roads... Does the TIGER data only have one category
for link roads?

It's also just struck me that none of the abbreviations in the TIGER data
have been expanded... i.e Avenue is Ave, Street is St, Drive is Dr, Highway
is Hwy, I'm guessing Pky is Parkway or something... This doesn't seem to
follow the standard of not using abbreviations in the data...

d
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20070906/96c71e9d/attachment.html>