>yes, armchair mapping generally bears the risk of misinterpreting the actual situation, (depending on what kind of tags you use this can be more or less important). Doing it in search and replace fashion is slightly different though, as you risk falsifying information someone else has meticulously surveyed (assuming that armchair mapping is mostly about the addition of new stuff).
and then you get the situation in Africa with HOT, new mappers less than 24 hours old, marking a tile done in HOT using tags which have extremely low usage in taginfo and are not the recommended tags for the project or area. It is tempting to do a search and replace for highway=pedestrian on the mapper and replace it with highway=path the recommended value in the African Highway wiki. http://wiki.openstreetmap.org/wiki/Highway_Tag_Africa I agree mapping done on the ground by locals is best. Cheerio John On 11 July 2016 at 09:45, Martin Koppenhoefer <[email protected]> wrote: > > > sent from a phone > > > Il giorno 11 lug 2016, alle ore 14:01, Éric Gillet < > [email protected]> ha scritto: > > > > I agree that survey are that on-premise survey is the best review > method. But then you are adressing armchair mapping as a whole and not > specifically search-and-replace edits. > > > yes, armchair mapping generally bears the risk of misinterpreting the > actual situation, (depending on what kind of tags you use this can be more > or less important). > Doing it in search and replace fashion is slightly different though, as > you risk falsifying information someone else has meticulously surveyed > (assuming that armchair mapping is mostly about the addition of new stuff). > > > > > > > There are instances where a survey is not necessary, and the information > needed to perform the modification is available on imagery, website or open > data, so the need to be on the premises is limited. But again this does > apply to all armchair mapping. > > > yes (let's say a survey is less necessary, still there is always the > possibility something has changed in the meantime and any edit will make it > appear as if the feature still existed at the time of the survey) > > cheers, > Martin > _______________________________________________ > talk mailing list > [email protected] > https://lists.openstreetmap.org/listinfo/talk >
_______________________________________________ talk mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk
