[OSM-dev] ways with 'spurs'

[Robert (Jamie) Munro]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Barry Crabtree wrote:
> I've come across some ways that go:
> 
>    A -> B -> C -> B -> D
> 
> Shouldn't they be done as:
> 
>    A -> B -> D, with a separate way B -> C?

IMHO, not if they're the same road, there's no point. I live in a
cul-de-sac with fork in it, and it would be stupid to have to tag it
twice, especially if this meant the name was rendered twice. Osmarender
and osm2pgsql (i.e. MapNik) both do the right thing. If you make sure
the first contiguous set of segments is the longest possible route,
renderers will write the name along that part of the way.

> On a more general note, is there any ongoing work to clean up the 'raw'
> database so there are no zero length segments, looping/branching
> segments. I know I've seen it mentioned a few times. If we can't rely on
> a clean starting point I can see loads of duplicate developments for
> handling these kind of cases....

It's probably high but not very hight on the list of priorities. Let's
get the rails port working first, then we can add validation as data is
entered, then we can fix the existing data and be confident problems
won't happen again.

Robert (Jamie) Munro
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFF4iG2z+aYVHdncI0RAjCBAJ4x4XOpeQsGuFigwKRi/SJvZ+slbACeOtc5
19vAKQh3nb/ERf+7YpuKmOs=
=lUdk
-----END PGP SIGNATURE-----