[OSM-talk] Visually detect missing roads

[Stephan Knauss]

On 15.09.2014 01:38, Dave F. wrote:
> On 14/09/2014 13:56, Stephan Knauss wrote:
>> First and most important: It can't tell you which data is better. It
>> just shows the differences.
> Then maybe I'm misunderstanding what you mean by: "So in a "perfect"
> area the map would be grey."
Hm, I think I understand. I should think of a better wording to explain 
it. "perfect" is not unambiguous so up to the interpretation of the 
user. It might also give the impression that Google map data is perfect.

What I tried to explain is a problem I encountered when I showed the 
diff first. People said: "why there is only a grey map in my area?"

So actually a map with no diff is good. At least a good indication that 
the map is not missing something important. Assuming for a moment that 
Google data is a perfect reference (which is not as we all know).

I have to think a bit longer on how to explain why there is "all grey".
Especially in central Europe and US there are not that many major 
highways actually missing.

East-Asia or Africa is a completely different story. Also Google 
accuracy varies much depending on their source.


> For clarification could you explain what you believe is inaccurate with
> this way: http://tinyurl.com/jwylzkb

The "South End Road" is changing it's type from a major highway into a 
residential just before "Bradfield Southend".

A usual tagging in OSM is to continue the tagging of interconnect roads 
as such even inside villages until it joins with another interconnect.
Compare it with "The Avenue" in "Chapel Row" a little to the west.

So "South End Road" should most likely continue as a major highway until 
it connects to "Common Hill".
Same "Marlston Road" in "Chapel Row". It should connect to "The Avenue".


As a side note: Is the spelling correct? There is a "Southend" village 
without space and in the road name it has a space.

> I genuinely believe we shouldn't be comparing OSM with Google. In so
> many ways the OSM database is far ahead of Google.
You are absolutely right with this. Probably every mapper can tell areas 
where there are serious flaws with Google data.


Stephan