[talk-au] Missing streets in Sydney

Anthony osm at inbox.org
Fri Sep 2 02:25:28 BST 2011


On Thu, Sep 1, 2011 at 8:40 PM, Ian Sergeant <inas66+osm at gmail.com> wrote:
> On 30 August 2011 16:41, Steve Bennett <stevagewp at gmail.com> wrote:
>> Surveying
>> suburban streets by GPS these days makes about as much sense as using
>> a horse and cart on a freeway...
>
> This tracing vs survey argument is as old as OSM is.  My vision of OSM is to
> get take a different route on the bike, or see more of a town when you are
> passing through, or even go for a walk around streets in your local area,
> rather than being a mechanical turk in front of a computer screen, but each
> to their own.

Personally I very much agree with this.  I'd never spend my time
tracing the roads of some boring suburb that I have no personal ties
to.  But I'm very glad that not everyone feels this way.

> Sometimes there is no alternative to tracing, but I think
> tracing without actually ever having placed a foot on the ground in the
> area, leads to a significantly poorer quality map, and you don't need to
> delve to far into the database for evidence of that..

Obviously a map is potentially better if one adds foot-on-the-ground
surveying to whatever other methods you are using.  But that's about
all one can say.

Tracing is quite often more accurate and/or precise than using a GPS.
If high res imagery is available, and it appears to be well aligned,
I'm pretty much always going to use that rather than GPS tracks, even
if I have done a foot-on-the-ground survey.

Put another way, unless your survey equipment is something equivalent
to a google car (http://www.flickr.com/photos/stewb2008/5840727837/)
or google bike (http://searchengineland.com/google-woos-brits-with-bike-based-street-view-project-19519),
foot/tire-on-the-ground surveying without using high res imagery also
invariably leads to a significantly poorer quality map.



More information about the Talk-au mailing list