The Scoop that is running kuro5hin right now (0.4.2 with some tweaks for performance-- still can't keep the server running. grrr) took about 2 months to emerge. If you peek back at this story, it would appear that Jan 9th was the first time Scoop did anything at all. At that time it could basically just display stories that had been preloaded into the database. So, today is March 8th, so yeah, almost exactly two months. [Insert gratuitous dig at how long it took slash to reach this level of functionality here ;-)]
About scaling: that is something I'm constantly thinking about. For example, the threshold to post stories is not a hard limit, but a percentage of registered users. Right now I think it's 3% (I adjust it depending on what voting traffic is like). So basically, say I had 300 users, a story would need a total positive score of 9 to be posted. It's still a threshold, but one that's recalculated to account for the number of potential voters. It seems to work well so far, although I'm considering improvements (like being able to vote for a rewrite, if you think it's a good story that needs some work, and potentially even collaborative community rewriting).
Other scaling issues are spam control and comment filtering. These will both improve. Here's the plan for spam control: Basically, I will choose a few regular contributors (who are willing) to assist in controlling spam. They will be able to "delete" comments, but when they delete a comment, it is not totally gone. It's simply moved to a special "admin-only" fake discussion. At any time, I can pop over there and see all the comments that have been designated spam by these people, and who did the deleting. If I disagree, I can reinstate the comment (and whack the deleter with a ruler or something). There will also be a function in there to nominate new spam-control officers. Any existing spam controller can vote that another reader should join them, and if someone gets enough votes, they join the team. This way, I hope, it will remain a job for "trusted" members of the community. Also, the potential for abuse is low, because every time they delete something, I *might* look at it. They'll never know whether I *will* see a particular comment, but I *could* see any of them. It's the classic panoptic system, which generally promotes self-disciple rather effectively. :-)
In concert with that, I may add a rating for spam, so that other readers can point out what they think should be removed. Perhaps the spam patrol wouldn't be allowed to remove a comment until it has been designated spam by X other readers. We'll see. I think there are ways to deal with scaling, without losing what makes the site special to begin with.
I'm also rewriting the code right now, to make it perform (*much*) better, and make it easier to add new features and whatnot. And there are some very cool things on the way, which hopefully will help the /. comparisons start to taper off. :-) kuro5hin will be getting less slash-like in the future. I took the slash interface as a model to start with, because it's nice to have a "spec" to code to. But at this point, I think I've done just about all the slash-ish things I want to do (except user prefs, but my preferences will be a lot more thorough). I expect the site to move into more uncharted territory soon.
The signal... well, I'm more interested in quality than quantity. Think of it as a signal quality ratio. If I'm interested and enlightened by 100% of the comments posted here, that is ideal. Whether there's one comment on a story, or 500. I think the signal quality on kuro5hin is hovering around 98% at this point, which pleases me to no end. I can only hope the quality stays where it is. Like I said, I would like to see more stories, but that will come.
Thanks for your questions. And dammit, I *still* haven't updated the FAQ. Too busy coding. ;-)
Not the real rusty
[ Parent ]