Re: [Talk-GB] �Correcting� existi ng data with OS Opendata

2010-07-23 Thread Simon Ward
On Tue, Jul 20, 2010 at 08:11:46AM +, Ed Avis wrote: > However I'd like to point out that 'ground survey' versus 'OS' is a false > dichotomy. Firstly the existing OSM topology is not always from ground > surveys. > More often than not it will be a trace from Yahoo aerial imagery. Secondly, >

Re: [Talk-GB] �Correcting� existi ng data with OS Opendata

2010-07-21 Thread Ed Avis
Simon Ward writes: >Other data has been shown to be quite inaccurate compared to OSM, >including mastermap data. Using our own collective knowledge just >produces better results anyway. I don't know what part of the country you are looking at, but in most cases I've seen, the OS data is so much

Re: [Talk-GB] �Correcting� existi ng data with OS Opendata

2010-07-20 Thread Ed Avis
Firstly I completely agree with the original point - unless the existing OSM streets are clearly unfinished work (perhaps indicated by being drawn with Potlatch but never given a street name, indicating no ground survey) it is not right to just replace them wholesale with tracing from OS. However