[Talk-GB] Retour de l'autoroute britannique

[Borbus]

On 15/12/11 01:22, Jason Cunningham wrote:
> I can see much of the woodland is an exact match for woodland is OSM
> whose source is OS vectormap district. So one of the sources is
> Ordnance Survey.

It certainly seems to be from Ordnance Survey.  But there are some
oddities.  Before the change, Google maps reproduced a number of errors
that had been revealed in OS OpenData by mappers.  When these are found,
the wrong names are tagged as not:name, you can see them for each area
here: http://www.itoworld.com/product/data/osm_analysis/main

I checked some of these errors after the change and lo and behold! they
are corrected, and now show the same as OSM.  However, other such errors
have arisen.

Now I have also checked to see if these exists in OS MasterMap.  From
what I can see, Google's current errors match OS MasterMap.  Before the
change they matched OS OpenData.  (Another question is how a different
set of errors to those found on MasterMap ended up in OS OpenData).

> There are several buildings, eg schools, with accurate outlines,
> where would this data have come from?

This is strange too.  I noticed that the new Gmaps data includes some
for the University of East Anglia: http://g.co/maps/u4pz3  But the data
found is beyond what is found in MasterMap, as that is simply the
building names.  It turns out they have exactly the information that is
found in the university's own PDF campus map:
http://www.uea.ac.uk/about/gettinghere/campusmap   I'm curious so
thinking about putting in a FOI request to the uni to see if they have
provided Google with this information.

-- 
Borbus.