[OSM-talk] Bing maps is misplaced

[john at jfeldredge.com]

In my case, I have done a mixture of image-tracing (from the Yahoo aerial imagery), and POI marking (from first-hand knowledge, and frequently from ccordinates measures using my phone's GPS).  All of my image-tracing has been in areas that I had first-hand knowledge of.  Since most of the roadways in this area are already marked from the TIGER import, and these imported roads mostly align with the Yahoo imagery, the majority of the image-tracing involves fixing the occasional road that was marked a few meters off from its actual location, while the adjoining roads are correct.  I have added a few minor roads that were built too recently to be in the TIGER data, plus mapping the zoo and a couple of small cemeteries.

-------Original Email-------
Subject :Re: [OSM-talk] Bing maps is misplaced
From  :mailto:lists at mail.atownsend.org.uk
Date  :Wed Dec 08 18:15:39 America/Chicago 2010


On 08/12/2010 21:59, Steve Bennett wrote:
> So the question arises: does the community support this view?
Unlike the Life of Brian, here everyone does seem to be an individual - 
I suspect that you'll get as many answers as there are mappers.

Speaking entirely personally, I do mostly only map places that I've been 
and don't tend to trace e.g. a road or track unless I've seen one end of 
it.  When I started adding stuff to OSM the map was entirely blank where 
I lived and I wasn't aware that anyone traced stuff at all.  I 
eventually encountered a road that seemed a bit wrong - it was 
consistently a few metres SW of my GPS traces.  Initially I assumed that 
I must be hitting some sort of "urban canyon" effect and tried again, 
but got the same results.  Eventually I figured that it had been traced 
from an old NPE (out of copyright) map.  There was actually nothing 
wrong with the tracing; the error was on the old map.

So was the original mapper wrong to have traced that road from NPE?  
Personally I'd say no; a road in not quite the right place (on an 
otherwise empty map) is better than no road at all.  Problems can 
obviously happen if what's being traced from isn't as good as it could 
be (the issue that Kenneth raised earlier on in the thread), and in the 
UK that may be an issue with some of the Bing imagery as it looks (a) 
quite detailed but (b) quite old.

Where there are a reasonable number of local mappers, tracing can be 
less beneficial because it can get people to think that an area is 
"complete" when it's not been ground surveyed.  When creating Garmin 
maps locally for my own use I try and incorporate certain "source="s 
(e.g. "NPE", "Yahoo") and users (if a known non-on-the-ground mapper) in 
the name.  I live not far from Staffordshire in the UK and parts of that 
are somewhat iffy - they look complete, but what's on the map doesn't 
match reality.  However, there are also places where a GPS simply won't 
get a good fix because of the terrain and short of a theodolite or 
something capable of dead reckoning, tracing is the only way to get an 
accurate road / path layout to lay POIs on.

If you're in a country with relatively few mappers, then tracing makes 
some sense, it gets coverage now where there otherwise would be none, 
but it doesn't take away the requirement for someone (eventually) to 
visit and add detail that you can only get by actually being there.  I 
don't think that there can ever be a planet-wide consensus about tracing 
where you haven't visited; but individual communities should be able to 
come to some sort of agreement for smaller areas.

Cheers,
Andy


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

-- 
John F. Eldredge -- john at jfeldredge.com
"Reserve your right to think, for even to think wrongly
is better than not to think at all." -- Hypatia of Alexandria