Alive... OR ARE THEY
Like many sites before them, these places indicate a sunny outlook, a clean bill of health and a total sense of "all systems go". But as we've found out from those many sites before them, fortunes can change overnight.
Archive Team considers these sites specifically of interest because they solicit so much content, contain so many works and projects by a wide group of people, or have the internet particularly dependent on them. Consider this a fire drill.. know what you can do to get your data off these sites and back them off for later.
- Academic Earth ( • • • ) has been worryingly unloved for a while, and holds a mountain of free education that's invaluable to the world.
- AnimeMusicVideos.org ( • • • ) is fine right now, but they rely on donations and host vast amounts of user-edited music videos on their server (presumably without mirrors). Hard to download as you have to be a member to get all the download links, and after downloading a handful you have to vode before you can d/l again (or you can donate which presumably gives you 1 year of free d/l access). Also, this site might be a grey area, copyright-wise, as the videos are all cut together from copyrighted material.
- Delicious ( • • • ) loves to change their API, which has a side effect of making it difficult to back up.
- Facebook ( • • • ) seems stable at the moment.
- FanFiction ( • • • ) represents many thousands of user-generated stories, essays and huge amounts of work.
- Friendfeed ( • • • ) is a happy clam who recently shacked up with Facebook.
- Google ( • • • ) wants you to think they will be here forever.
- Infoanarchy ( • • • ) The site is functioning again. Might be worth backing up, though. For months, a simple database error that could be fixed with one command KO'd this site unexpectedly with a wealth of P2P information lost.  The site was down for four days in June 2010. There is now an archive of the content at: infoAnarchy wiki archive
- Internet Archive ( • • • ) seems stable at the moment but its 2 petabytes of data aren't mirrored anywhere else, the code for their system isn't open source and generally they're a single point of failure for a large amount of the web's history. Why should there be only 1 internet archive?
- There seems to be a second instance at Bibliotheca Alexandrina although it's currently broken and out of date.
- Know Your Meme ( • • • ) is at this point the de facto central repository for information on internet memes and culture. It is as popular as ever at the moment, but even with this popularity, former owners Rocketboom had trouble financing it. In the spring of 2011 was sold to Cheezburger Networks, a site which has been known to "reorganize" its properties, sometimes with a detrimental effect on content. Though it was quite a different story, I might remind people what happened to Encyclopedia Dramatica.
- Last.fm ( • • • ) is being cloned by free software developers in the form of Libre.fm -- they have a tool, Lastscrape which can get all your listening data out into a tab delimited text file.
- Literotica.com ( • • • ) Contains over 290,000 user-written stories and poems. First pass at a backup: part1.rar, part2.rar, part3.rar, part4.rar -- contains the text of all stories as of the backup date in XML format. (One page of one story is missing because it doesn't exist on the site; embedded images and audio are not included this time; non-English stories aren't labelled with their language).
- LiveJournal ( • • • ) fired a bunch of US-based developers, but is still serving from its new (presumably cheaper) data center in Montana.
- Pixiv ( • • • ) and deviantArt ( • • • ) are the largest Japanese and American (respectively) fanart (and valuable art in general) collections on the internet.
- Pouet ( • • • ) is an important site of the demoscene. It indexes and ranks demoscene productions ('prods') and also includes a free-for-all BBS-style forum.
- Reddit ( • • • ) is where many of the users have now migrated. Stable for now, but team is small.
- SourceForge ( • • • ) is a critical repository of open source code, information, and webpages. It is mirrored and maintained, but there are sure to be parts that are neither.
- TVTropes ( • • • ) is a popular wiki dedicated to finding recurring patterns in fiction, and discussing fiction in general. No word on whether there are backups. The administrators have a tendency to delete things indiscriminately, usually to save on disk space: article edit histories are frequently purged, and old forum threads have been known to get deleted mercilessly. A backup is now available.
- Twitter ( • • • ) is tweaking away.
- WebCite ( • • • ) itself seems to be having trouble with funding, and is facing "possible discontinuation." As this site serves as a stable reference for fleeting Web references, it would be pretty disastrous if it went away.
- whitehouse.gov ( • • • ) is up and running for #44,
but we've lost all info for #43. (See also: kottke and Read Write Web.)and #43 is available at http://georgewbush-whitehouse.archives.gov/ thanks to the Presidential Records Act. We also want to watch out for site changes / disappeared pages that were embarassing or whatnot.
- Wikia ( • • • ), the for-pay arm of Wikipedia (just kidding, it's a different company, but shares a lot of people) is a repository of directed, unsubject-to-wikipolitics wikis, many of them intense and completist. It'd be bad for them to go away.
- WikiLeaks ( • • • ) contains several thousand leaked documents from sources such as the Iraq War and the cables famously known under the label 'Cablegate'. Due to the content on the website, and that PayPal and Amazon (very) quickly dropped their hosting for them during Cablegate's opening days, it should be considered a potential target for any number of government committees for quick shutdown.
- Wikipedia ( • • • ) will surely be here forever and ever! Fortunately, we don't have to take their word for it as they offer dumps of the data minus the photos. However no-one has verified that Wikipedia can actually be restored from these dumps. If disaster strikes then we could discover a serious problem.