On Wednesday 19 October 2016, James wrote: > Ok, Christoph now that I am at a proper computer I can address your > comments: > [...]
Thanks for the additional information. Based on this i modify my comments, what i am still missing is: - link to the source data for the buildings - information on attribute and geometry processing - this refers to how the source data is processed to generate the OSM files. Based on the address data this is clearly non-trivial and needs further explanation (expansion of abbreviations, conversion of upper case names). - information on what you actually intend to import - please forgive me for not taking the time to scrape every data tile from data.osmcanada.ca to determine the extent of the data. Additional notes: - the address data is clearly not ready for import in the current form, it contains lots of duplicate addresses and positions seem pretty far off in some cases, often too far to be properly matched with buildings without on-the-ground inspection. - the building geometries contain various flaws like poor orthogonalization, overlaps and impossible geometries like here: https://www.openstreetmap.org/#map=19/45.32737/-75.70919 and systematic errors like here: https://www.openstreetmap.org/#map=19/45.32802/-75.71617 - it is unclear how you intend to determine which building geometry is more accurate if there is already data in OSM. > If you are just going to skim over documentation, what is the point > of creating it? The purpose is to document the import to allow mappers to look up how this particular import is done. If finding out a specific detail about it - like for example to determine how the value of a certain tag is determined or which area exactly is affected by the import - requires you to read the full text from top to bottom, it does not fulfill this purpose that well. -- Christoph Hormann http://www.imagico.de/ _______________________________________________ Imports mailing list [email protected] https://lists.openstreetmap.org/listinfo/imports
