"Active following network of trust spam filter"
"Censorship and shadow bans"
Same thing, yes or no?
Discussion
Very difficult, if not impossible, for the first not to cause the second.
The path to hell is paved with good intentions...
If you can't turn it off
This is the key. We should have a simple setting and not require an entire fork to disable it.
First of all: 👍
Second: I am really annoyed by a mob using the word censorship because the people they follow and trust reported someone. Don't follow people who abuse this, it's their credibility on the line. The accounts are only getting a warning for you, not censored or shadow banned on nostr or on amethyst for everyone. People really get confused when they hear that as it is not close to the same thing as on Twitter. It abuses the word.
Third: I'm not saying it's perfect the way it is implemented now, I also think that user should have more options, but the PR was removing the feature entirely instead of making it an option. So is the fork. I have no problem with people using it, but I don't think using aggressive language about a spam filter being censorship is helping anyone and even damages nostr.
I agree with you. Gotta stay balanced. I do see both sides especially since spam filtering is far from standardised in Nostr yet.
We're in agreement it should be an option, not forced on or off.
The problem wrt content being flagged is you don't see "this has been reported by these people" anymore, the message just doesn't show, and certain accounts like nostr:npub1nym2200088t397yx43lza4gfec3rk56gxnwjnk2u4w5ld0qpau6sskc7fq get auto blocked... that is shadowbanning when you can't disable it.
Nym is not blocked for me. Check the reports tab and see who reported. See who of the people you follow did.
Not blocked for me either because I had to unblock him. But try this: make a new keypair, look up nym from it in Amethyst, see what happens.
I agree this should not happen. Especially not when not following any of the reporters.
Exactly. And that's not even following the "web of trust" model. It's just straight up shadowbanning.
I think spam is a difficult issue as it's subjective. Using nostr:npub1nym2200088t397yx43lza4gfec3rk56gxnwjnk2u4w5ld0qpau6sskc7fq as an example, my guess is he got caught in the amethyst spam filter twice. 1 he regularly posts the same "what's for supper ..." message which I believe nostr:npub1gcxzte5zlkncx26j68ez60fzkvtkm9e0vrwdcvsjakxf9mu9qewqlfnj5z has said amethyst blocks accounts with the same post multiple times. 2 at one point he was zapping 1 sat to every post which some people complained about and likely blocked him for. While I don't consider either of these spam and actually like to see them, I can see how some may call that spam. Which is why I think the current amethyst spam filter while well intentioned without being able to opt out is too far. I respect Vitor's position that software should be opinionated, and that open source allows it to be forked. In relatively new ecosystem like this I think it's healthy and in the long run will lead to the best software and user experience.
I think you hit the nail on the head. And it demonstrates the difficulty in general that any software has in balancing out blocking clear cut obvious spam (e.g. those Optimism messages) and genuine users whose activity happens to trigger an algorithm.
And all fairness to nostr:npub1gcxzte5zlkncx26j68ez60fzkvtkm9e0vrwdcvsjakxf9mu9qewqlfnj5z I've spoken to him about why one of my accounts, nostr:npub1ms4z8e2fsgjh75vtk8c9pym50gjvp3uxl7uru8v7c7pnd9empadq2ulevr, got caught up as spam in Amethyst and he was very transparent about it and helped me remedy the issue (which I've since done - it used to post duplicates, but now I've updated the code so it has to run through the full list once before it starts again, and put that code on GitHub).
He's definitely taken the fairest approach possible, although I'm not sure why nostr:npub1nym2200088t397yx43lza4gfec3rk56gxnwjnk2u4w5ld0qpau6sskc7fq is still blocked when many on Nostr enjoy the excellent AI art he shares.