JA Tech-Tips: Twitter’s API is changing in March – what you need to know
Twitter’s API (application programming interface) has become so popular that it’s rare you don’t see a website that has a feed of Twitter content or a way to login using your Twitter account. In August 2012, Twitter announced some upcoming changes to its rules and regulations to foster a more consistent user experience across platforms and devices. The deadline to meet the requirements of API v1.1 is March 5, 2013.
This could affect any website that uses the Twitter API and many third-party apps, such as Hootsuite or Flipboard, who have had to make major adjustments to adhere to Twitter’s new rules.
Even Journalism Accelerator has had to make some changes. Are you ready?
In this article we will discuss the three changes that we’ve encountered at the JA. Full details can be found in Twitter’s Developer Rules of the Road. We encourage you to check out what this may mean for your site, to help you anticipate how to make this transition with minimal stress or site disruption.
Twitter’s new display guidelines for developers are now being referred to as requirements and must be followed in order to use their API. This may interfere with the unique styling on some websites, but the intention is to maintain the same look-and-feel and functionality of twitter wherever you encounter it. That means being able to see a user’s profile information like avatar, name, username, and timestamp, and being able to take actions like reply, retweet or favorite. To make this change a little easier, Twitter has created a simple interface to embed a single tweet or to build your own timeline for Tweets from a user, a user’s favorites, lists, any search query or any hashtag. They generate a snippet of code ready to copy and paste.
Working directly with Twitter
The current API allows any “entity” (person, company, app) on the web to make API queries. Twitter in some cases has no idea who, or what, is making the request. With the new API, an “entity” making an API query must be authenticated. In other words Twitter now knows who and what is making requests. This allows Twitter to put thresholds and limits in place on certain levels of “entities” so they can better manage entities that are abusing the free service with an extremely high volume of automated requests. They have stated that they’re not shutting down applications that currently exceed their new limits and they are willing to work with clients directly if they need to make a large number of requests.
Dropping RSS/ATOM/XML support
It seems like every site has a Twitter feed, but Twitter insists the use of RSS is still low, and as of March 5, they’re dropping support. If you’re using an RSS feed to display your tweets on your website, you may want to switch to a timeline. Many aggregation sites and apps have used RSS as a simple way to pull tweets into other services.
What has the JA had to change?
We’ve had to make some very specific updates to our Tweetcatcher code to use Twitter’s new API v1.1. We are in the final stages of the Tweetcatcher conversion. The JA’s Tweetcatcher is the element you see on the right hand side of most all JA pages with tweets on a rolling display. Many of JA’s content specific pages (resource listings or blogs or questions) have custom searches set up to only pull tweets related to that item. We’ve had to create and register our “application” with Twitter and have an API key that we now use to enable JA to pull tweets for our custom searches. The last step is squashing one final display bug. As currently planned, we are on track to meet the March 5 deadline.