• OpenStars@piefed.social
    link
    fedilink
    English
    arrow-up
    1
    ·
    13 hours ago

    For the image issue, if it is a link then wouldn’t Lemmy detect that already? Links to news articles and such at least work that way but I don’t know about images.

    • Martineski@lemmy.dbzer0.com
      link
      fedilink
      English
      arrow-up
      2
      ·
      edit-2
      6 hours ago

      This is a link detection and not an image one. Two wildly different things. It also wouldn’t handle images with slight differences like an edit here and there because again, doesn’t handle images. Same goes for varying levels of compression. In fact it wouldn’t even detect the exact same images with different sources or when reuploaded by users. Even if there were people who source images from the same place it would still be irrelevant without an overwhelming share of the users doing that to make the feature actually relevant. And EVEN if there was this high coordination then any trackers, shorteners, arguments, etc. varying the link to the same source they would be treated as a different links without recognising them as a duplicate like with youtube for example. So users would need to be a literal mindhive to coordinate on this level and at this point the tools would be pointless because the knowledge would be shared between everyone anyway.

      Having this feature would help immensely both as a poster and as a mod to handle the images with high probability of being a repost. But at the same time I know it isn’t feasible due to image processing requiring quite a bit of computing power so it will continue to be a dream.