Eight months after I requested my own Twitter data from Twitter through a legal request under the European privacy law, Twitter now allows you to download your own tweets through their interface. The archive can be downloaded from the settings page (see this blog post from Twitter) and the file named tweets.zip contains all your tweets from the beginning.
The tweets are stored in two different formats: CSV and JSON which makes it a versatile archive to work with for both users and developers. The archive does not only contain your own tweets but also tweets you have retweeted but excludes DMs and favorites. The archive is neatly organized and tweets are stored in files per year per month, for example: 2007_08.js. The .zip file also includes an interface to browse through your archive per year per month:
My previous archive which I received from Twitter contains more data because back then I requested all data Twitter keeps about me, which includes direct messages, metadata and logins, IP addresses, contacts, etc. The data that is available per tweet in both archives is quite similar:
When comparing my old archive to the new archive what seems to be different however is the availability of a retweet count. The old archive contained a line “retweet_count”: *, which would show the number of retweets for that particular tweet. This (valuable) data has been removed from the new archive.