4 ms·

See https://news.ycombinator.com/item?id=13593334 https://news.ycombinator.com/item?id=13593334

• boulos 10y ago
  And I should add: if any of you work at The Met or the Rijksmuseum and want to be in touch about us hosting a CC0 copy of the images, please get in touch! (We've tried to reach out via the emails we could find, but I don't think we've received a response yet).
  • toomuchtodo 10y ago
    While I think its very cool Google would reach out to host a copy, the canonical copy belongs in the Internet Archive. EDIT: No snark intended.
    • boulos 10y ago
      A canonical copy certainly does, but just as folks don't want to keep scraping the Met, I'd prefer that folks who want to do style analysis or color transfer or whatever else, do so with a public dataset in GCS (and S3!).
      • toomuchtodo 10y ago
        I appreciate your efforts to provide such a copy :) Apologies if my parent comment sounded negative, it was not mean to be! A rising tide lifts all boats.
        • boulos 10y ago
          Not at all! I'd argue it's extremely clear that if there could be only one allowed copy, it'd be with the Internet archive (and failing that Wikipedia? It's not even the same ballpark, but I think that's second). I've just come around to the "more copies are better" standpoint, especially for different needs.