About a year after I complained about Twitter’s lack of archiving, they implemented a kind of backup solution involving a zipped tweet archive.
The Twitter archive is a bit of a headscratcher. It’s a searchable offline snapshot of tweets stored in JSON files organized into a year/month navigation UI — usable in any browser, but you must first request the archive from Twitter Settings and then retrieve it as a ZIP file after the server builds it and notifies you by email.
The nice part is that your full Twitter history comes to you as a portable web app that you own; a self-contained snapshot to store or upload to your preferred web host. On the down side, why decouple the archive from the stream in such a manner that you must return to the download page for a new snapshot every time you want an updated version? It seems like Twitter could have deployed the archive as a feature of every user profile, with the downloadable archive as backup. It seems a disjointed approach to what should be a simple archiving problem, which makes me wonder what kind of coding issues Twitter faces that would make a dated message log anything other than straightforward.
For now, my tweet archive lives in the history section, where I also keep backups of the embarassing older iterations of my web presence.