Pieren wrote:

> I'm currently implementing the cadastre support in JOSM for the french
> part of the island "Saint-Martin" shared with our Dutch friends (it is
> a special projection).
> The island is quite well mapped today, mostly from the hi-res Yahoo
> imagery I guess:
> http://www.openstreetmap.org/?lat=18.069&lon=-63.0746&zoom=13&layers=B000FTF
> 
> The problem is that the data, although they match the Yahoo imagery,
> seem to be shifted from about 800 meters in north.
> The French national geographic institut provides files about geodesic
> reference points and one is marked on the main fortress here:
> http://geodesie.ign.fr/fiche_geodesie_OM.asp?num_site=9712701&X=491000&Y=1998000
> Here the details of one mark on the ground:
> http://geodesie.ign.fr/fiche_point_OM.asp?num_site=9712701&no_ptg=01
> 
> So, the fortress should be at 18.0707416944 lat and -63.0851921944 lon
> but in OSM it is at (approx.) 18.0705758642 lat and -63.0845414733 lon

I'm not quite clear how you come to that conclusion. While JOSM only displays 4 
digits after the comma (ooh, can we change that to 7 someday?), both 
coordinates 
lie in a wooded area which may very well be the fortress.
And the difference between these two points is 71 metres in total, just 18 
metres N-S difference, not 800.
The Google images are a lot better than the Yahoo ones, and where Google says 
the fort is, is close to the first set of coordinates, and that lines up with 
Yahoo. But in OSM, there is no point to identify the fortress.

Furthermore: I downloaded the GPS data for Sint-Maarten and that lines up very 
nice with the roads of OSM (both in the French as the Dutch part), and the 
roads 
line up with the Yahoo imagery.

So I don't really see this mismatch you are seeing.

Regards,
Maarten

_______________________________________________
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Reply via email to