[Talk-GB] Road Names Quarterly Project
tony.wroblewski at gmail.com
Tue Feb 17 11:45:41 UTC 2015
I've seen numerous examples of this in OS OpenData, including fields
marked up as woodland (which aren't on their paid paper maps), also
they move things out of the way, e.g. buildings and drains to make way
for the roads. Also, if anyone has used the OS VectorData to any
extent, they'll notice it's also been altered (simplified)
unnecessarily, or random things have been omitted for no apparent
reason, e.g. large ponds, whilst the small ones are included. It
almost seems to me that the OpenData is purposely created this way to
limit its usefulness (e.g. No vector version of StreetView, missing
tracks (but they include historical roman roads??), so they've done
the absolute minimum to make their data open.
Compared with other countries who offer all of their data for public
use, it's pretty sad unfortunately. I think it's safe to say that OS
OpenData should not be used as an authoritative source.
On 17 February 2015 at 11:31, Chris Hill <osm at raggedred.net> wrote:
> Local authorities name streets. If the name board on a street is wrong
> people tend to complain and get it replaced. If the street name on an open
> version of OS data is wrong no one complains. I would always trust the name
> board rather than OS open data. Strangely enough the names on the streets
> match the names on the paid-for versions of OS data (that we can't use in
> OSM) so I wonder how and why the OS open data has the names wrong in the
> first place.
> Cheers, Chris
> On 17 February 2015 10:03:42 GMT, Colin Smale <colin.smale at xs4all.nl> wrote:
>> It's only "correct" because that's the frame of reference you have chosen
>> in this case. The local authority decides what a street is officially
>> called. How that is transposed to signs sometimes introduces errors, and
>> these errors are sometimes volatile. The OS is not the source of the
>> official name either is it?
>> On 2015-02-17 10:45, Philip Barnes wrote:
>> On Mon Feb 16 23:35:41 2015 GMT, Pmailkeey . wrote:
>> On 16 February 2015 at 15:51, Philip Barnes <phil at trigpoint.me.uk> wrote:
>> In these cases you should check the name on the signs and if osm is wrong
>> correct it. I my experience osm is often right and os opendata is incorrect,
>> in these cases add the opendata name to a not:name tag. Where osm in
>> incorrect it is often caused by an awkward spelling, so a photo can be
>> useful. Sommerfeld Road in Telford took me a few attempts to get right,
>> originally mapped as Summerfield. Phil (trigpoint)
>> What's the general consensus where *current *OS data and the sign is wrong
>> ? Should OSM show the wrong name but flag it as being wrong or show the
>> correct name and add the wrong name as a not:name ?
>> OSM should show the correct name, which is the one on the sign.
>> Not:name is there to suppress the error and to indicate a mapper has
>> surveyed it and shown the OS are incorrect.
>> From my experience OSM is usually correct, the biggest cause of error is
>> OS getting apostrophes wrong, and if OSM is wrong it is usually a weird
>> spelling where the mapper has remembered the name but has forgotten how it
>> was spelt.
>> Phil (trigpoint )
>> Talk-GB mailing list
>> Talk-GB at openstreetmap.org
> cheers, Chris
> osm user, chillly
> Talk-GB mailing list
> Talk-GB at openstreetmap.org
More information about the Talk-GB