We (your moderators) have been mulling over an idea for a good way to deal with old, unanswered and unanswerable posts. We've been in discussion with the StackExchange team and while we're a long way from actually launching this, we feel it's about time we involve the community and let you help shape how such an event would work.
###So what would happen in Clean-Up week?
Our working plan is to form a list of all the questions that aren't acceptably finished. They will be of a certain age and will either have no answers or answers without enough votes to count as a real answer. With that list users will be able to do one of several actions to help clean up the site and at the end of the week, the users who have been most helpful will get a reward.
The aim is two-fold: we heavily decrease the number of dead questions and get more users involved in the community both at an answering level and in maintenance.
The actions for dealing with a thread are pretty similar to new, active questions:
Identifying good answers - Plenty of old questions do have some answers but those answers haven't been voted for nor have they been accepted. In these threads we need people to vote for the good answers, just like normal. Getting a good answer up to +2 will classify the question as resolved and get it off our list.
Answer it - You probably haven't seen every question on the site so some of these questions will get organic answers. You'll get rewarded by the community in the regular way.
Flagging old topics - Given the nature of Ubuntu releases we need to make sure we prune out unanswered old topics about old installation techniques or hardware issues that may be fixed by upgrading Ubuntu. This won't always be automatic but we need to keep on top of how many active, old-release questions we have.
Removing or improving the unanswerable - Ask Ubuntu is only as good as its questions. We will need people to chase up old threads to make sure they have enough information to be answered. If their users don't respond and edit their posts, we need to cull the deadwood.
Merging duplicates - So many duplicates slip through the gaps and we end up with multiple answered questions covering identical ground. While this isn't the prime aim of the week, any duplicates (even from established, answered questions) are candidates for treatment.
Identifying good, hard questions - Some questions are unanswered because they're really very tough to solve. There's no value in closing them so if you can't answer them, we need users to try and involve relevant external groups and get askers to keep the information flowing in. If we can match people with the knowledge with the questions, we fulfil several of our goals.
###Rewarding participation
As part of the discussion leading up to this post, we've been talking about rewarding users who jump in with both feet. There are still some technical issues to rectify that SO Inc might not be able to help us with though, so if you're reading this and you know lots about the SE API or relevant systems development, we may need your help in getting a system set up to handle things.
The main problem is keeping track of the list of target threads (which we will generate for the week) and what actions people do on them. We need to track new answers, comments and flags (even after moderators deal with the flag in question) so that when we reach the end of the week, we can see who has done best.
As for the rewards, we're negotiating with SO Inc to see what we can do. Extra points for anything but answering questions appears to be out of the question but physical "swag" (Ubuntu t-shirts, mugs, mice, etc) might be an option depending on the geography.
###When will this week be?
TBA. We need to overcome technical issues to track the progress through the week and until we can do that, we can't really hold a competitive event like this. Ideally this would occur prior to the 11.04 Natty Narwhal release.
###Any ideas, questions, etc?
This is still a very young idea so we'd really like your input especially about the process of doing the week where we're still technically trying to work out how we do it.
