[OSM-talk] Comparing Google and OSM

Rick Collins gnuarm.2006 at arius.com
Sun Oct 21 03:08:36 BST 2007


At 08:19 PM 10/20/2007, you wrote:
>Rick Collins schrieb:
> > I am curious.  What makes you think your GPSr is not accurate?  Do you
> > see clear differences between your unit and others?  What model do you
> > have?
>
>It's an eTrex yellow, 1st genration. Trackpoints are set by very crude
>sheme and the GPX is useless without the lines drwan between each point.

Yes, a friend has one of these and I find mine is much better, even 
though it is 4 or 5 years old.  His also goes to pot when under tree 
cover or even rain.  Current units work well even indoors.


>And it's inaccurateness can be easily spotted by driving through a road
>twice and doing some rectangular movement in between. In the suburbs,
>with broad streets and low buildings the offset isn't usually more than
>10-20m. But downtown with multistorey-houses right behind the streetedge
>it's offten beyond fifty meters.

Multipath comes from the many reflections in urban environments and 
is a problem for all current generation GPSrs.  I expect them to be 
dealing with multipath in 2 to 3 years.  It has bubbled close to the 
top of issues now that so many other things have been dealt 
with.  The DSP technology is just about there if they can pull it off 
without greatly increasing the power consumption of the front end.


>That are days with good satellite constellation, but I've also spend
>afternoons with creating totally useless tracks because the offset was
>bigger (and constantly changing) than the gap between two parallel streets.
>
>That's why I was not surprised when seeing a difference between my and
>Google's data.

Yes, it is time for a new receiver.  Is US$100 a significant 
investment for you?  I paid US$45 for my current unit on eBay.  That 
was cheaper than my bluetooth receiver that I paid US$70 for, and is 
only usable with a PDA.

Rick





More information about the talk mailing list