There's been some discussion of how to improve the structure of LessWrong at the site software level - for example adding subreddits or modifying how main and discussion work. One roadblock to this that's been mentioned is a shortage of programmer hours. I'd like to volunteer mine.
I recently finished a course on web development in which, among other things, I build a Reddit clone using Ruby on Rails and Backbone.js. It's been several months since I've written any Python, and I'm somewhat wary of the time required to get familiar with the LessWrong codebase, but I think think the time would be worth it for me: it could potentially improve LessWrong a lot and would let me tick off my "have contributed to an open source project" box.
Of course, before any of that happens, there needs to be some agreement on what changes we think would be a good idea. So... discuss.
EDIT: For context, it's been suggested that part of the benefit of subforums is it could defuse debates over "what topics are appropriate for LessWrong." We could even have an "off-topic" subforum, a common feature of online discussion forums - I think bringing the format of LessWrong more into line with what's standard on other websites could help newbies be less confused here.
https://code.google.com/p/support/issues/detail?id=1210 suggests sorting by stars should be possible.
Personally, I wouldn't waste any time on a discussion of a potential big fix unless the instigating party had already shown the ability and willingness to hack on the codebase by, for example, fixing a number of 'little things'. (Since usually such people never contribute anything and so discussing their proposals is a waste of time and energy.)
Fair enough, and I see the link for sorting by stars is here.
Should I literally just be going down the list top to bottom? Or limit myself to things that are accepted / authorized?