[Moderation] Revert request - Russia, Irkutsk - Evgeny Mandrikov

[Dave F.]

Peter Miller wrote:
> On 2 Oct 2009, at 19:17, Dave F. wrote:
>
>   
>> Peter Miller wrote:
>>     
>>> This list is about two issues really, one is developing the required
>>> technical tools to be able to spot and deal with vandalism and
>>> mistakes and determine if something is vandalism or what. The other  
>>> is
>>> to sort out the social protocols to deal with issues that come up, so
>>> lets use this case as test case!
>>>
>>> There is firstly an issue about the import being badly done, then
>>> there is a possible copyright violation.
>>>
>>> With regard to the first I guess we could have tools that monitor the
>>> minutely feed and spot possible dodgy data coming in, such as area
>>> features that are not areas. We might wish to ignore them if there  
>>> are
>>> only a few, or possible come in with a 'mentoring' hat on and help  
>>> the
>>> person, or with some fix-up tools to correct the errors, or we may
>>> decide to reverse out the changeset and talk about it with the
>>> contributor before they try again. All of the above will be
>>> appropriate at times. Clearly we don't have the in place now.
>>>
>>>       
>> To self check I use:
>> http://keepright.ipax.at/report_map.php?db=osm_EU&zoom=14&lat=48.20808&lon=16.37221
>>
>> Maybe sites like these should be listed in the newbie welcome email?
>> Are there others sites like this?
>>     
>>> Then there is the issue of potential copyright violation. The best  
>>> way
>>> I can think of to do that is to overlay the data on TeleAtlas data,
>>> Navteq data and any other official data sources you can think of and
>>> see if the fit is to good to be true.
>>>       
>> I believe "to good to be true" is subjective & could lead to  
>> penalising
>> excellent, accurate mapping by an editor.
>>     
>
> If there are roads shown through the middle of houses visible on  
> google satelite view on both the data imported into OSM and on another  
> source and the roads line up then I would consider it appropriate to  
> take a very good look at the reality. If the reality is that a road  
> has not replaced a house (or vice-versa) then I is about as good  
> evidence as you could get for infringement
>   
OK, but that's a pretty blatant example, more of a "too /bad /to be true"

Let's take a winding, non navigable stream. The OSM way is very near, 
but not quite replicating the Google underlay.
Is that worth questioning? How accurate would it have to be before it 
raise alarm bells?

Sorry if I appear pedantic, but I think it's important to clarify the 
boundaries of what's classed as a violation.

Cheers
Dave F.