With the Minecraft release in the near-future, would it be worth it to systematically go through all the old questions (say, 621 of them) when games are officially released and make sure they are still relevant?
This is a problem with any game that receives updates to a point, but Minecraft has undergone myriad radical changes since early beta so there are bound to be posts (questions and answers) that are now totally irrelevant and could confuse new users (to the game or our site). Presumably after an official release the amount of changes will slow, so doing a formal review then, if ever, would make the most sense.
Old questions about mechanics that no longer exist could be closed as localized and maybe a little disclaimer added to the top (as per here), which could include a link to a newer, related question (if one exists). Maybe also retag as below.
Answers that are obsolete seem a more difficult problem, as they've been upvoted and/or accepted and will totally bury the newer, up-to-date (read: correct) answers. This guideline seems totally idealistic and hasn't really been followed. Users that are still around could be coaxed into editing their prior answers or accepting different ones, but there's no guarantee they would, or if they do, put the same amount of effort into it that (presumably) earned them their top spot in the answer ranking. If the askers or answerers aren't around and a revised answer requires a bit of work, it seems a tad unfair to edit someone else's answer and give them those reps.
If the question is pretty much the same, but the mechanics are totally different, could the question be duplicated (re-asked) and change the tag on the old one to gamename-beta? Stack Overflow seems to have plenty of questions that are pretty much identical save for (as an example) being Python 2.x, or 3.x, etc, though it's far less likely that anyone will still use an old version of a game when the new one comes out.