4 ms·

Yes, if I had improvements I was confident about, I'd list them. (Maybe don't think of this as mainly a news site? Older stories are popular.) The first idea to

Yes, if I had improvements I was confident about, I'd list them. (Maybe don't think of this as mainly a news site? Older stories are popular.) The first idea to come to mind was to slow down the aging and make the front page into a sample of the stories with upvote support over this newly longer interval -- but I have no experience managing a site like this. But "quick, upvote before it's gone" seems clearly the wrong policy to encourage in people.

hyperpallium 10y ago
Non-news: historically, hackernews was startup news, which then expanded to "gratifying intellectual curiosity." I'd like to split the "new" and the "interesting", rather like how newspapers have a magazine-like section on weekends. Some of the most popular articles are quite old as you note. eg Feynmann stuff. Your idea would work well for this. But I doubt anything so drastic will be trialled... though dang seems a lot more willing to experiment than pg was, so maybe it's possible, if we can think of an effective yet low-cost way to test it? BTW there is a hidden lower-gravity version, https://news.ycombinator.com/best https://news.ycombinator.com/best (but of course most votes aren't made using it, so it doesn't test the idea). I agree blind upvoting is a bad idea, though I read the plea as "if you are already a fan of our work, please upvote early". In practice, the super-short story commented ("the little gods") is probably the best way to go - a quick way to evaluate the offering (which incidentally is what I did before upvoting).
- abecedarius 10y ago
  Making the front page change on reloading or for different readers (making it a random sample) would indeed be really disruptive, but it could have a kind of second-order benefit, that new experiments would become easier to introduce, because we'd stop expecting it to look the same for everyone. Not that I'm seriously advocating this for HN -- maybe for a new site? Geez, I forgot about /best. I tend to visit here to shirk work, and I'm not even shirking most effectively. That's a great suggestion, pointing out the short story. I enjoyed it too.

hyperpallium 10y ago

Non-news: historically, hackernews was startup news, which then expanded to "gratifying intellectual curiosity." I'd like to split the "new" and the "interesting", rather like how newspapers have a magazine-like section on weekends. Some of the most popular articles are quite old as you note. eg Feynmann stuff. Your idea would work well for this. But I doubt anything so drastic will be trialled... though dang seems a lot more willing to experiment than pg was, so maybe it's possible, if we can think of an effective yet low-cost way to test it? BTW there is a hidden lower-gravity version, https://news.ycombinator.com/best https://news.ycombinator.com/best (but of course most votes aren't made using it, so it doesn't test the idea). I agree blind upvoting is a bad idea, though I read the plea as "if you are already a fan of our work, please upvote early". In practice, the super-short story commented ("the little gods") is probably the best way to go - a quick way to evaluate the offering (which incidentally is what I did before upvoting).

abecedarius 10y ago
Making the front page change on reloading or for different readers (making it a random sample) would indeed be really disruptive, but it could have a kind of second-order benefit, that new experiments would become easier to introduce, because we'd stop expecting it to look the same for everyone. Not that I'm seriously advocating this for HN -- maybe for a new site? Geez, I forgot about /best. I tend to visit here to shirk work, and I'm not even shirking most effectively. That's a great suggestion, pointing out the short story. I enjoyed it too.