Hacker News new | ask | show | jobs
by jerf 3772 days ago
I've made a minor hobby of studying how communities are formed by the technical structure of the community, and what interactions they enable. The control of visibility is actually a subtle art, and a lot of things like Twitter or Facebook start with an initial design that is quite unsubtle that works at first, but scales poorly.

I'm not convinced Twitter is fixable as-is. The most obvious technical solution is to do what Reddit did and make sub-Twitters, but they're not going to do that. But as one big global broadcast platform, it too easily exceeds the ability of the human mind to deal with things when you get even modestly popular, to say nothing of being a celebrity trying to directly participate.

But in reality there's simply no way to be a member of a "community" of that size. Arguably that's one of the places where this essay sort of stalls out... there's no community here to be discussing in the first place. Just a really, really big pile of people, with affiliations too loose to even remotely be a community. If there is "power" here, it's not at all clear to me who has it, and it certainly seems on the evidence that Stephen Fry is actually on the short end of that stick rather than the long one. To the extent that the essay seems to vaguely try to suggest that he did the wrong thing, itself ironic in light of, well, itself, I don't think the case was made.

2 comments

Even sub-communites don't solve the problem. There are certain subreddits notorious for dogpiling on posts in other subreddits. The article mentions this indirectly by observing that what starts out on one platform can easily migrate across communities, even to communities you're not aware of.
I can't say I've made a study of it, but I've certainly thought about it.

I'd be interested in your thoughts on reddit style voting with visible points and reordering. In my observation each of those things are detrimental, at least with respect to community building and maintenance.

It seems to scale to medium size OK (several thousand lurkers, ~100 regular posters and a steady stream of drive-bys), though I observe even a medium sized subreddit doesn't feel all that different than a Usenet forum of the same size. Voting may let the community get a little larger than just a threaded conversation, especially because it can distribute spam filtering, but I'm not sure it's that meaningfully different in the end.

I'm also not convinced it's terribly detrimental. Most of the accusations against vote systems are true of threaded forums in general. Group think develops, regardless. People can be hounded out of a community, regardless, for the same set of reasons. It seems to me the real key is the use of a threading system in the first place, rather than the votes attached to the threads.

It may be fair to say that threaded voting enables the largest communities I know about. At the scale they break down, I don't know anything that works. I've never seen a mechanism for having a tightly-knit single community of larger size without being what most of us would consider a community fail.

By contrast, threaded vs. flat is a vital change.