Simon Poole wrote Maybe I'm misunderstanding what you want to do, but it would seem as if that just amounts to doing the work twice. After revising 126 elements, I agree with Simon. Let's avoid doing the job twice. Below is an evaluation for Quebec / Canada with a sample of 126 of 1,316 id's from Frederik list. I see that we cannot assure in this sample what is the name source for 98% of the ways modified.
Since there are often references to Canvec, I did an Overpass query. It appears that 53 ot the 1,316 have the tag "canvec:UUID" and this tag was sometimes modified. But without confirmation from chdr it is difficult to assess if it did really come from Canvec. Then, yes, for Canada at least where we have opendata available, let's simply redact all the names except the typos that are easily identiable and start over. 2 typo corrections (ie. place to Place, St- to Saint-) 110 Name added 14 Name revision (it could be adding Street, Avenue, Rue, Place, Croissant, etc) I saw one Google artefact where Rue was modified to Avenue while Rue is specified on the OpenData source regard Pierre
_______________________________________________ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk