Why are third parties important in the Twitter ecosystem?

Let Twitterrific count the ways:

First use of “tweet” to describe an update (see page 86 of Dom Sagolla’s book.) First use of a bird icon. First native client on Macintosh. First character counter as you type. First to support replies and conversations (in collaboration with Twitter engineering.) First native client on iPhone. And more.

http://furbo.org/2011/03/11/twitterrific-firsts/

Expanded for ecosystem firsts beyond Twitterrific:

http://blog.140proof.com/post/30593225507/twitter-features-i...

PhantomJS is great for scrapping, but the proposed solution can still be easily recognized as a bot behaviour. Twitter can easily block it for excessive page loads, for example serving CAPTCHA the same way Google does to prevent search scrappers. I can't see how this solution would be more solid and future-proof.

Completely right. A lot of social sites, notably Facebook do this when you for example post bit.ly links, so users are already groomed to accept it.

Then again, you could buy external captcha solving services and keep on trucking. :)

I'm sure it will work great, until they alter their page layout ever so slightly. Change an id here or there. Then you have to scramble to fix for the new layout while your 'API' is down.

Needless to say, this is an awful idea for anything besides a small personal project. PhantomJS or no, it's easy to detect this behavior, and also easy to understand why Twitter would be against it.

Twitter has been around for its first 5 years without putting a single advert on their site, and only NOW they want to make profits after realising the third party developers are hosting Twitter based content elsewhere with mobile client apps?

That is a totally bad business decision from the start if you ask me.

You've not seen the Promoted Trends and Promoted Tweets in Search, that have been there for nearly 2 years. They've been profitable, on a monthly basis for a while.

A big problem with this is from my own experiences, despite the rate limiting on the API, it has far more access to historical data (eg, past tweets) than the browser.