Hacker News Reader

Comment by cousin_it | Hacker News Reader

I agree that pull is the best model for posts. For comments I still think the best model is to push them to the post author, and let the post author moderate. This way people who can behave civilly can get an initial audience by writing comments, and having a link to their blog in the profile, with no intermediaries involved. If someone doesn't like the moderation, they're always free to write a comment-as-post, where they can write anything they want but have to take care of distribution themselves.

View on HN

What if someone posts a dangerous lie? The post author is going to delete any comments that expose the lie. How would comments-as-posts be sufficiently linked to the original post so the lie can properly be exposed to as much of the audience as possible?

What if somebody posts a dangerous lie on a centralized platform and blocks everybody who criticizes it? Even if it doesn't allow deleting comments, you can effectively the same result with just block.

A mute-only social media platform would be doable technically but it would likely involve tons of spammers replying to large accounts. Think about how unusable the replies to an American politician's tweets are on Twitter because the US courts have ruled that it is unconstitutional for politicians to block people on social media.

The best solution to this problem I've seen so far is Community Notes on Twitter where a crowdsourced fact check is directly pinned to the tweet allowing users to challenge liars without directly calling them out and getting blocked. More centralized approaches to fact checking don't work particularly well because you end up with biased fact checkers who clearly have an agenda and thus aren't trusted at all by the liar's audience.

the thing is, with no intermediaries involved, more specifically, without any proper "search engines" (or "aggregators") involved, the network will suffer greatly from a content discovery problem regardless (as the fediverse currently is, IMO).

with presence of such services, the problem of comments (and reactions in general) can be solved too. if a poster is ok with engaging with potentially hostile content, then they can get reactions to their posts from aggregators that aren't heavy handed on moderation. if they don't want to bother with such interactions, then they can choose safer aggregators. if they want, they can only pull reactions from feeds they are already subscribed to, similar to private posts on twitter.

I'm more thinking about the social side of things. When comments follow the pull model, you get the pingback problem of old: most "comments" will be links to blog posts, themselves stuffed with more promotional links and so on. The only way to avoid it and have comments look like a somewhat nice garden is to allow post authors to say: please write text comments and don't stuff them with links, or you won't pass moderation on my blog. In other words, the push model. In my experience that's the best solution to this particular problem.

I think that's an issue of tools at your disposal to create content (including comments) rather than their distribution model (e.g. pull vs push). If your main method of saying stuff is through a blogpost, ofc you'll end up in the situation you've described. If the tools at everyone's disposal are textboxes that just work, people will use that.

Yeah, this is precisely what I said as well, being able to comment on anything and have that seen even without a follower count is important for making initial connections on a social network.

Reply on news.ycombinator.com