On 26/05/2015, Martin Koppenhoefer <dieterdre...@gmail.com> wrote: > My main concern with wikidata for the moment: it's mostly as fuzzy as > Wikipedia is - because the objects are not created by humans but conversions > of articles. Using only wikidata would mean we are sure that wikidata will > be a success.
Agreed. I initially tought that wikidata was mostly human-curated, which led me to believe that it could only play catch-up with wikipedia in terms of content and that wikidata items would not always be available for OSM tags. The fact that pedia and data are kept so tightly in sync (one being a metadata backend for the other) is a mixed blessing. It's nice that you can always count on a wikipedia item existing if the pedia article exists, but it adds a lot of noise. The combination of stable ids and tracking non-stable articles means that a lot of bloat must accumulate. Seeing that there's 13.9M wikidata items for 4.8M English wikipedia articles doesn't reasure me (I know it's not all bloat, but still). _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
