Canary in a coalmine

Twitter API changes are a death sentence to its ecosystem

Elliot Bentley
twitter-locked

It’s the end of the line for many Twitter app developers, but celeb-crazed users are unlikely to care.

It’s difficult to avoid today’s headlines on the upcoming
changes to Twitter’s API, announced via a blog
post
. Unlike
previous threats
, these changes aren’t just hints as to where
the service will go, but a death sentence to many parts of the
Twitter ecosystem.

Let’s get the less controversial bits out of the way first:
changing the way in which Twitter limits API requests – 60 of each
endpoint per hour, rather than 350 of any per hour, is sensible,
and mandatory OAuth use will help reduce load, even if some apps
will suffer.

Meanwhile, the strict guidelines
on tweet presentation may seem sensible on the surface – except
they seem to even extend to contexts such as the redesigned
digg.com, which uses custom embedded
tweets in its current design.

Generating more public ire is Twitter’s decision to limit
third-party clients to a maximum of 100,000 users without explicit
permission (or, for those already above 100,000, the ceiling will
be double their current userbase).

Mind you, the shutting-out of third-party clients has been a long
time coming. Twitter’s Director of Consumer Products, Michael
Sippey writes in today’s blog post:

Nearly eighteen months ago, we gave developers guidance that
they should not build client apps that mimic or reproduce the
mainstream Twitter consumer client experience.” And to reiterate
what I wrote in my last post, that guidance continues to apply
today.

What Twitter wants, as illustrated by a rather cringeworthy
chart, are analytic apps and corporate clients.

The current API, version 1.0, will be discontinued in six months to
give developers time to adjust. Of course, developers in many cases
won’t be adjusting – they’ll simply be abandoning their projects.
Instapaper developer Marco Arment highlights in a blog post
just how many services are likely to be affected – including his
own app.

As you might expect, the reaction to all of these announcements has
been largely negative. John Gruber, a big fan of third-party
clients, is scathing of Twitter’s direction:

So Klout, which is utter vainglorious masturbatory nonsense,
that’s OK. But services like Storify and Favstar, which are actually useful and/or
fun, those are no good.

Some have played down the announcement, accepting that Twitter
is taking fairly sensible steps. Anil Dash’s
rewrite
of the official blog post puts a cheery spin on things,
but his interpretation of Twitter policy – that third-party clients
are still welcome, and that apps with more than 100,000 users will
get “personal” treatment – is probably a bit naive.

But this bad press is nothing to Twitter, because they don’t
need its core tech audience anymore. An audience of ‘Little
Monsters’ and ‘Beliebers’ (that’s Lady Gaga and Justin Bieber fans)
is infinitely larger and more profitable.

App.net’s alpha client

Alarm bells have been ringing for a while, and developers have
been looking for an alternative service for a while. The
announcement of API v1.1 comes soon after the funding of App.net, a
service pitching itself as a Twitter clone. The difference here is
that it is paid for by users rather than its advertisers, and with
an unrestricted API. While still currently a proof-of-concept
“alpha prototype”,
App.net has succeeded in giving itself initial momentum. Where the
service is heading is another matter, and the likelihood is that it
will end up being a closed community of Twitter-bashing techies
willing to pay.

Twitter was of course initially populated by early adopters, many
of them developers interested in playing with the service’s API.
But App.net is unlikely to ever gain Twitter’s mass appeal: after
all, what does it have to offer users who don’t even know what an
API is? A greater range of clients and no adverts are no substitute
for Twitter’s millions of users (notably their celebrity
contingent). Plus, in an age where most services are free and
ad-supported, $50/year for a single account is a hard sell.

The other social network that has failed to take advantage of
developer dissatisfaction is Google+. A year on from launch and
still a full API is nowhere to be seen. While it’s to be expected
that, like Twitter and Facebook, Google want to keep control of
their product, opening it up to outside developers could give the
service the shot in the arm it needs to become somewhat relevant.
Unfortunately, this doesn’t look likely.

Facebook has proven that “mainstream” users already invested in a
service are very, very difficult to scare off, and so far Twitter
hasn’t offended them directly. Let’s just hope they can keep up a
high standard of native apps.

Author
Comments
comments powered by Disqus