Last spring an announcement from the platform team at Twitter not-at-all-subtly suggested developers of third party clients should find something else to do, and today a list of changes to its API turns that whisper into a firm nudge. The limit that most directly affects any of the unofficial clients you may be interested in using is that existing apps currently servicing more than 100,000 individual user tokens will be allowed to double their current count, but cannot add any users past that without Twitter's permission. Going forward, any app that needs more than 100,000 tokens to do things like access the timeline, show DMs or anything else a client app might do will also need Twitter's permission to operate. Other changes include that any pre-installed client app on something like a phone, computer, or TV will need Twitter's permission before it ships (sensing a trend here?), or potentially face revocation of its application key. Moving on, the Display Guidelines about the information any app that displays tweets must provide are shifting to Display Requirements, with violators potentially losing that application key. Those Twitter Cards that started rolling out over the last few months are also getting a big push, with plans to include other ways for developers to bring their rich content to Twitter, and embed real-time Twitter content on existing websites.

In a section of the blog post that specifically calls out popular third party clients like Tweetbot and Echofon, it puts them in a zone of Twitter apps that it believes developers "should not build" since they mimic or reproduce the mainstream Twitter client experience. Other services focused on aggregation like Storify and Favstar.fm are also called out as not having much of a future in Twitter's view of its universe A-OK, see the update below. It's pretty clear where we're headed, as Twitter seeks to monetize a growing and talkative user base, it wants to make sure users are all seeing things in the format it wants them to. With a six month migration period ahead, developers and users may have hard choices to make. Tweetbot developer Paul Haddad has already tweeted that "the sky is not falling...the cap is pretty huge and we aren't going anywhere", and we'll undoubtedly hear from others soon. The rest of the details reside beyond the source link, but let us know first -- are you learning to love the official Twitter clients, or are you thinking paying $50 a year for an entirely unproven alternative with no users doesn't sound so ridiculous after all?

Update: Twitter platform director Ryan Sarver tweets that Favstar.fm and Storify are actually "good examples" of services it would like to see. Also, TweetLanes developer Chris Lacy is apparently encouraged by the change, thanking Twitter for "giving client devs a chance"