Re: [Tagging] bridge AND embankment

2015-07-23 Thread Eric Sibert
@Eric: I looked at more examples, and I have to admit that you are right with your statistical (0.1%) argument. Most cases I looked at, are obvious accidental tagging errors. I checked for Madagascar. I found one case and I'm the author :-p I first added embankment and later cute a small part of

Re: [Tagging] bridge AND embankment

2015-07-23 Thread Volker Schmidt
@Martin: you are right there, naturally. I should have said "You cannot do this automatically with a mass edit. You need to consider each situation. In many cases you may be able to understand the situation from satellite photos (for example from shadows!) but there will be a large number where you

Re: [Tagging] bridge AND embankment

2015-07-23 Thread Martin Koppenhoefer
sent from a phone > Am 22.07.2015 um 22:16 schrieb Volker Schmidt : > > Which means manual correction with inspection of aerial photos. I agree on all but this sentence: you won't be able to actually "inspect " a spot on aerial imagery, rather you should go there to understand what is at th

Re: [Tagging] bridge AND embankment

2015-07-22 Thread Eric SIBERT
Although I agree that such combination is suspicious... > 250 in France A rough evaluation for France give me 200k ways with bridge=yes. So about 1 error each 1000 bridges. Not such a big issue. My 0,02 €. Eric ___ Tagging mailing list Tagging@op

[Tagging] bridge AND embankment

2015-07-22 Thread Volker Schmidt
By chance I have detected that amny mappers use this combination on highways: ways con bridge=* embankment=yes To see an example search in Northern Italy: http://overpass-turbo.eu/s/axW When I saw the first example I thought of an accidental mapping error, but the Overpass Turbo revealed that thi