[Imports] Florida Landuse Import
Hiausirg
grussausbw at gmx.de
Thu Mar 31 17:42:40 UTC 2022
"you'll find many things conflated with coastline ways"
True - but of course, i didn't suggest to do this blindly everywhere. People should judge for themselfes if they want to replace coastlines or use the existing ones. Most coastlines merged with other features are in urban areas - which aren't touched anyway.
"It wouldn't astonish me to find a great many sections of coastline where the mapping has been improved since the original import."
Also true - if they are detailed enough (many from grouper are, for example), they should be kept, or at least their history by using JOSMs "Replace Geometry" feature. But still, many coastlines are completely untouched and need a lot of cleanup.
A few words on the tasking manager approach: I don't think it makes sense to create 60 individual projects for 60 counties. Maybe for the first 1 or 2 counties, but not for everything. Creating 5 projects by using the borders of the 5 water management districts is way easier, and the data quality won't suffer either.
"I don't think that it is a bad thing for an import to take a long time. If time is needed to do it well, then things will take longer."
True! But how well is well enough? I recently deleted large parts of a 13-year old landuse import in Georgia, because everything was so incredibly horrible quality. (https://www.openstreetmap.org/changeset/119065140). No one bothered to do this, in 13 years. It wasn't even hard. Or other example from FL: Someone adds landuse=forest to a large protected area boundary, and certainly not everything was "landuse=forest". It stays for over a year, until i removed it recently (relation 3499712). No one else cared. And everything here is supposed to be absolutely great now, the last JOSM validator error where two landuses overlap a half inch or whatever has to be fixed? I don't get that.
More information about the Imports
mailing list