2 ms·
I've been playing with this for the last 20 minutes, and there are things I really like and things I don't like. The good: It's got a lot of the automatic thin
by jeremysmyth 14y ago
I've been playing with this for the last 20 minutes, and there are things I really like and things I don't like.
The good: It's got a lot of the automatic things that make Stack Exchange a pleasure to use - conversations slide into place nicely, infinite scrolling feels nice and new, and updates to conversations happen while you're watching.
The bad: The front page is already very noisy, and it's only in test mode. I expect that with time, the only way to use this properly will involve creating "channels" with tagging or filtering.
There are two major problems with this outcome. Firstly, if users select their own "channels", it becomes a reinforcing cycle where each "channel" (or "room") is only exposed to its own conversation. This is largely what happened with USENET (and to a degree what happens with subreddits), and while each one might be good if it stays small, if it doesn't it'll end up being as noisy as the front page. If managed well, on the other hand, I expect that the prettiness of conversations as they are now will follow nicely into each channel.
The second problem with the noisy front page is that as with every other general purpose discussion site with a front page, there will be a race to the bottom, where everything that makes it to the front page will be about grumpy cats or hot girls.
Maybe my criticisms stem from the very nature of discussion forums (look at the cycle of slashdot, digg, reddit etc.), but I don't see this tech fixing that problem like Stack Overflow claims to have solved the Q&A problem. I'd like to think it will though.