4 ms·

The direction seems fruitful, but there was very little to learn in this article. More insight into what is actually wrong with hosting providers would have mad

The direction seems fruitful, but there was very little to learn in this article. More insight into what is actually wrong with hosting providers would have made this a lot better. I now feel about Hacker News round about how I felt with TechCrunch a few months before I stopped reading it completely. I'm mainly here because of inertia, and because there is still great content from time to time and I don't want to miss it. My attention is very much up for the taking. The moment something better comes along, I'd be the first to switch. I wish PG put more love into HN. It used to be amazing, now it's just nice. At least, that's my opinion.

swader 14y ago
Perhaps you can help us all turn Quibb (http://quibb.com http://quibb.com) into something like that?
- davidkatz 14y ago
  Quibb seems interesting, but from your homepage it's hard for me to get a sense of value. May I ask why you don't let me see your content?
  - swader 14y ago
    I'm just a user, but the gist behind the "restrictiveness" is weeding out the chaff and avoiding the flood of quantity over quality that's overly present on App.net and other networks and content aggregators. The entry policy will be much more open later on, once it comes out of alpha/beta/whatever state, but for now, a personal contact with the founder and CEO, Sandi, is all that's required, just so she gets a sense of you being a real person who knows what they're talking about. Would you like an invite link so you can see if the content is (or can become) to your liking? Anyone who would like one should feel free to ping me via geekli.st, (http://geekli.st/Swader http://geekli.st/Swader)
    - davidkatz 14y ago
      At least the homepage doesn't seem that restrictive to me. It looks like it was designed to communicate something about Quibb and get me to apply. I just think it does that very well.
grep2 14y ago
What would be required to grab your attention? Curated content? A better discussion platform? Or just plain more signal per noise? One approach I'd like to see implemented is some monte carlo comment display algorithm. Assign every user a base point level (based on his contributions/selectable?), show every comment/post with score above this level, but comments below with decreasing probability. This has the advantage of allowing new/initially low-ranked stuff to get picked up faster than a hard threshold, while keeping the noise at a level acceptable for the user. Or is this nonsense?
- davidkatz 14y ago
  Sounds like a good direction, like most things it probably needs to be tested in the wild. I'm not sure that it addresses the heart of the problem though, which might be this: as communities get larger, quality drops. This is a hard problem to solve, and all I can do is offer some ideas. One idea might be to have a hierarchy of influence, where some users have more power than others, and users get promoted automatically according to their activity. Off the top of my head, here's one possible way this could work: Say you have three member tiers - reader, contributor and editor, and let's further say that the power of each tier differs in order of magnitude. readers can grant items 1 point, contributors 10 points, and administrators 100 points. If members at a higher tier than you consistently upvote stuff after you upvote it, at some point you get promoted. In the same way, if they downvote stuff after you upvote it you get demoted. I'd be interested if HN or Reddit ever experimented with something like this. One other important thought - you'll probably never be able to completely automate a great community. Or put another way - everything else being equal, a community that is also actively curated by someone with great taste is probably going to be better.

swader 14y ago

Perhaps you can help us all turn Quibb (http://quibb.com http://quibb.com) into something like that?

davidkatz 14y ago
Quibb seems interesting, but from your homepage it's hard for me to get a sense of value. May I ask why you don't let me see your content?
- swader 14y ago
  I'm just a user, but the gist behind the "restrictiveness" is weeding out the chaff and avoiding the flood of quantity over quality that's overly present on App.net and other networks and content aggregators. The entry policy will be much more open later on, once it comes out of alpha/beta/whatever state, but for now, a personal contact with the founder and CEO, Sandi, is all that's required, just so she gets a sense of you being a real person who knows what they're talking about. Would you like an invite link so you can see if the content is (or can become) to your liking? Anyone who would like one should feel free to ping me via geekli.st, (http://geekli.st/Swader http://geekli.st/Swader)
  - davidkatz 14y ago
    At least the homepage doesn't seem that restrictive to me. It looks like it was designed to communicate something about Quibb and get me to apply. I just think it does that very well.

grep2 14y ago

What would be required to grab your attention? Curated content? A better discussion platform? Or just plain more signal per noise? One approach I'd like to see implemented is some monte carlo comment display algorithm. Assign every user a base point level (based on his contributions/selectable?), show every comment/post with score above this level, but comments below with decreasing probability. This has the advantage of allowing new/initially low-ranked stuff to get picked up faster than a hard threshold, while keeping the noise at a level acceptable for the user. Or is this nonsense?

davidkatz 14y ago
Sounds like a good direction, like most things it probably needs to be tested in the wild. I'm not sure that it addresses the heart of the problem though, which might be this: as communities get larger, quality drops. This is a hard problem to solve, and all I can do is offer some ideas. One idea might be to have a hierarchy of influence, where some users have more power than others, and users get promoted automatically according to their activity. Off the top of my head, here's one possible way this could work: Say you have three member tiers - reader, contributor and editor, and let's further say that the power of each tier differs in order of magnitude. readers can grant items 1 point, contributors 10 points, and administrators 100 points. If members at a higher tier than you consistently upvote stuff after you upvote it, at some point you get promoted. In the same way, if they downvote stuff after you upvote it you get demoted. I'd be interested if HN or Reddit ever experimented with something like this. One other important thought - you'll probably never be able to completely automate a great community. Or put another way - everything else being equal, a community that is also actively curated by someone with great taste is probably going to be better.