3 ms·
I don't disagree, but that's just one particular scenario. But on platforms like blocking a news source, because they posted a spammy article 3 years ago can e
by winternett 4y ago
I don't disagree, but that's just one particular scenario.
But on platforms like blocking a news source, because they posted a spammy article 3 years ago can end up preventing you from seeing the valid posts they make now after firing their editor.
I am not at all against blocking features, but there are logical ways to better manage these features, like perhaps creating an additional view/page where you can view recent posts from accounts and people you've blocked (with controls to unblock them embedded next to posts).
Another possible solution to the blocking conflict is to be able to completely block posts based on keywords in addition to being able to block user accounts... Blocking users should be a nuclear option rather than the only option available, and it should work consistently across the platform.
- notacoward 4y agoAgreed. There are better blocking/muting/filtering systems, and worse ones. Implementations also vary widely. For example, I think the design of filtering on Tumblr is perfectly fine, including keywords, but the implementation is garbage and stuff from blocked users "leaks through" in reblogs. It would be great if there were a widely agreed-upon set of principles (preferably not a formal standard) for how such features should work, but I'm not aware of any. Then again, it's very far from what has ever been my professional specialty so maybe I'm just missing something.