[Talk-us] Night of the Living Maps

Richard Welty rwelty at averillpark.net
Fri Oct 26 01:46:13 GMT 2012


On 10/25/12 9:14 PM, Mike N wrote:
> On 10/25/2012 12:34 PM, Martijn van Exel wrote:
>>   We're
>> talking about promoting a US project like fixing TIGER deserts / Ghost
>> towns[1] for the next NOTLM (date to be set). Good idea?
>
>   I would say that this is a good idea.   If a curious person (non-OSM 
> contributor) were to look at a garbled set of TIGER streets, they 
> might not even start with basic contributions such as verifying the 
> streets where they live and work, or adding POI.   Learning how to 
> correct those TIGER streets would be a huge obstacle.
>
>   So my input would be to find not only TIGER deserts, but those which 
> appear to have been created from GPSs carried before the GPS signals 
> were unscrambled.   Generally, these were entire counties.   Some of 
> the TIGER datasets have been updated to current technology standards. 
> Detecting this could be by comparing the alignment of unedited OSM to 
> latest TIGER.
>
>   Otherwise, I'm not sure how to tell if a TIGER desert was good to 
> begin with, or just never touched as well as being far out of alignment.
i don't think that any TIGER desert is really trustworthy. i've done 
TIGER review of much of the Capital District
of New York, and while street alignment is usually tolerable, there's a 
persistent error rate, mostly things
that are only really detectable by a ground survey.

however, there are vast areas (like most if not all of WV) where the 
alignment of TIGER 2005 is really
awful, and TIGER 2011/2012 is a significant improvement. since the 
latter is available as an image
layer which i know works with JOSM and i understand works with Potlatch, 
it's pretty reasonable to
load up a spot in the desert and compare unedited data with 2011/2012 
and realign. just don't remove
the reviewed tag, that shouldn't be done without a ground survey.

richard




More information about the Talk-us mailing list