On Thu, Mar 14, 2019 at 4:37 PM Mateusz Konieczny <[email protected]> wrote: > I would advise to tag just forest landcover is satellite images are unusable > to tag > other features properly and to not introduce incompatible tagging scheme just > because > you really want to vectorize this specific low quality dataset.
I think there is some grain of value here. He has a use case where he has valid but incomplete information: "in this polygon, I don't know what the landcover is in sufficientdetail to tag it with the existing OSM taxonomy, but it is definitely NOT forest." Otherwise, there's no way to distinguish, "we haven't read and mapped the imagery yet" from "we've mapped the imagery, and there's no forest here." When I know something about a piece of ground, I think it's reasonable for me to want to tag what I know, without needing to do further investigation. That's why, for instance, I object to distinguishing power=line from power=minor_line by voltage - I may have no way of determining that in the field, but the power line is still a landmark, and still deserves to be mapped. I'm not demanding that for the specific use case of 'NOT forest', that we have to have a broadly accepted tag, but I think that it may well be a valid reason to invoke "Any tags you like" https://wiki.openstreetmap.org/wiki/Any_tags_you_like. It's really offensive to tell mappers that they can't map features because we can't figure out how to tag them, and in many cases it comes across as cultural insensitivity: "the data model is fine. Fix your country." _______________________________________________ Tagging mailing list [email protected] https://lists.openstreetmap.org/listinfo/tagging
