Edit: obligatory explanation (thanks mods for squaring me away)…

What you see via the UI isn’t “all that exists”. Unlike Reddit, where everything is a black box, there are a lot more eyeballs who can see “under the hood”. Any instance admin, proper or rogue, gets a ton of information that users won’t normally see. The attached example demonstrates that while users will only see upvote/downvote tallies, admins can see who actually performed those actions.

Edit: To clarify, not just YOUR instance admin gets this info. This is ANY instance admin across the Fediverse.

  • Popsip@pawb.socialEnglish
    211·
    1 year ago

    So if one downvotes someone and then removes that vote, does doing that removes it saying they downvoted or does it still keep it on record?

    • Muddybulldog@mylemmy.winOPEnglish
      21·
      1 year ago

      I had to run an experiment on this one.

      It appears that changing you vote causes the old vote to be completely deleted from the database and a new vote cast and propagated.

      Edit: The above description is what happens in the COMMENT_LIKE or POST_LIKE table HOWEVER the ACTIVITY table reflects both actions, which makes sense since it’s a complete transaction log. So, it’s a slightly more complex query but the history is maintained.

      • dack@sh.itjust.worksEnglish
        10·
        1 year ago

        However, that’s not really any better for privacy. There’s absolutely nothing preventing someone from logging a history of the changes.

        • Muddybulldog@mylemmy.winOPEnglish
          4·
          1 year ago

          Upon further experimentation, the ACTIVITY table maintains both transactions so… there you have it.

    • Funderpants @lemmy.caEnglish
      93·
      1 year ago

      Depends on the rest of the structure of those tables and the supporting procedures that modify them. I haven’t checked, but I’m very interested in using this as a sample dataset.

        • Funderpants @lemmy.caEnglish
          9·
          1 year ago

          I guess, but it’s less “I don’t know” and more “it’s not knowable” from the screenshot alone. We need to see more of the schema to answer the question. “I don’t know” fails to communicate what would be needed to answer the question, it’s an okay answer, but I think mine is more useful.