[OSM-newbies] Street names and Nominatim

[john whelan]

Ta for the info, another theory down the drain, I wonder if its to do with a
place name.  No I haven't touched Richmond road but I thought it had been
labeled street on one segment.  Merkley has been there for some time.

Thanks

Cheerio John

On 24 May 2010 21:54, Richard Weait <richard at weait.com> wrote:

> On Mon, May 24, 2010 at 9:34 PM, john whelan <jwhelan0112 at gmail.com>
> wrote:
> > What does Nominatim search for a street name?  street=xyz or name=xyz?
> >
> > It matters because in Ottawa most roads are labeled name=xyz not street,
> and
> > Nominatim doesn't seem to find them but the odd one or two that are
> labeled
> > street=xyz do come up.
> >
> > Example Richmond road is labeled street=richmond road, Nominatim finds it
> > whilst Merkley Drive, Ottawa is not found.
>
> I'm not familiar with street= as an OSM tag.  osmdoc.com shows 37 uses
> of street= and 29 more of Street=.   I suggest that using street as a
> tag is probably a misunderstanding and street= is unlikely to have
> support in many tools.
>
> addr:street is used more than 3 million times and is part of the
> Karslruhe Schema for addressing.  So the error may be related to
> addressing.
>
> http://wiki.openstreetmap.org/wiki/Proposed_features/House_numbers/Karlsruhe_Schema
>
> name= is used over 16 million times for road names and many other
> object names.
>
> Nominatim is still reporting an out of date database.  It is expected
> to be repaired some time in May.
> http://wiki.openstreetmap.org/wiki/Nominatim
>
> Once repaired, I would expect nominatim to find Merkley Drive normally.
>
> I did find a Richmond Road, but it was already name=Richmond Road.
> did you fix it?
>
> Best regards,
> Richard
>
> _______________________________________________
> newbies mailing list
> newbies at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/newbies
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/newbies/attachments/20100525/e4e083bd/attachment.html>