[OSM-talk] Visually detect missing roads

[Stephan Knauss]

Salut Sylvain,

On 16.09.2014 14:14, Sylvain Maillard wrote:
> I look at your map for Lyon, and don't understand most of the "missing
> road" that your tool is showing ...
> an example with
> http://compare.osm-tools.org/?zoom=15&lat=45.73417&lon=4.82971&layers=BT00F
> : the road is there in both maps and seems to have the same kind of
> attributes (oneway and classification). Can you explain why there is a
> big red mark on it ?
When you have a closer look at the data you notice that in OSM the "Rue 
Marcel Mérieux" is classified as highway=residential. This is a minor 
road classification.
In Google the road is listed as a major highway.

As this is a difference it is marked as red (=different).

I can't tell you which one is correct. This needs to be determined by 
the local mappers. There are different tagging schemes and ideas of how 
roads should be classified inside towns. Often this also differs 
depending on the country.

I got feedback from mappers who use the tool in central Europe to detect 
possible tagging problems. But my focus for the tool was to detect real 
"white spots" on the map. Areas where major highways are completely missing.

The best use of the compare tool is outside of big cities. Have a look 
at rural areas in Africa or Asia to get an impression on how useful it 
can be to complete the major road network there.

Or French Guiana
http://compare.osm-tools.org/?zoom=15&lat=45.73417&lon=4.82971&layers=BT00F

Some areas might be more difficult to process than others. Consult Bing 
or alternative backgrounds like mapbox what could be a correct mapping. 
If you can't figure it out with a reasonable certainty, leave it as a 
job for future on the ground mappers or waiting for better imagery.

Stephan