- cross-posted to:
- [email protected]
- [email protected]
- cross-posted to:
- [email protected]
- [email protected]
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.
Which is my concern. I don’t like Reddit having and selling that data, but it’s easier for me to trust-ish a singular entity than some entire web of random people, which probably includes some corporate people siphoning data anyway. I know some would likely find that a tad paradoxical, but that’s how my brain works. At least then the corporation can be held accountable per the standards of the region they’re based in should there be issues, or users can mass target the corporation rather than go “Don’t like it, just move to another instance.”.
For reference, it’s still not ideal, but I’d somewhat trust my instance’s admin. Why can’t my vote history be shared purely with them? Then give other admins the raw upvote/downvote data of the post/comment. After all, the instance I choose my account to be on is my decision.
It’s not. I am careful about what I put online. Whilst I’m uncertain as I’ve never particularly tried to do so beyond some cursory Googling, I’m pretty sure you can’t tie my username back to me IRL. But even so, there’s no need to add to the pile of potentially traceable publically available data.
That can still be anonymised behind a hashed ID. If all my votes were registed to some User-XXXX and it wasn’t possible to retrieve my username from that, I’d have no issues. Though from my discussion with other people, it seems that’s counter to how ActivityPub intrinsically works. I’m increasingly working towards the opinion that the fediverse isn’t for me, if it’s all set up in a similar fashion and apparently unchangeable. As they say, “different strokes for different folks” I guess.
You can’t do this part. It makes it way too easy to just say “This post, -1000. This (shill) post, +1000.” Having to put names to those thousand votes makes a difference. A hash really doesn’t, as a hash isn’t hard to fake. The other solution is like mastodon, where your votes only count on your own instance. That decision would basically kill small instances of Lemmy, so I can understand why they didn’t go that direction.
I don’t know if you’ve ever noticed the difference between Reddit’s Hot and New, but it’s extremely dramatic. Votes are important, and that makes it hard to effectively not use them on smaller instances.
So yeah, I decided to look into ActivityPub. From what I’m reading, it seems like the sacrifices in privacy are an intentional decision by the creators of the protocol so that admins can weed out “undesired interaction”.
I can certainly see where they’re coming from, and I’ll be interested to see how it plays out. But ultimately, I don’t like this philosophy for a Reddit-like site, so sadly I don’t feel comfortable enough to contribute to it any longer. I guess it’s my fault for not looking into it before signing up, but what can ya do.
Regardless, thanks for the discussion, to you and everyone else. Hope you guys do well here.
Thanks, and good luck. The only parting thought is that if you don’t want the public to have that data (and you may have a point), I wouldn’t feel comfortable giving it to Meta or Twitter either.