Click here to preview the new Fast Company

Want to try out the new

If you’d like to return to the previous design, click the yellow button on the lower left corner.

Think You're Pissed at Facebook and Twitter?

This morning you logged on to Facebook and Twitter, only to find you had no way of announcing to the world that you were starting a new diet. Or eating Raisin Bran. Or hungover. So you stormed away from your computer, irate: how could the intertubes have failed you so utterly?

If you think the combined stuttering of Twitter, Facebook, and LiveJournal this morning—which has been attributed to several DDOS attacks—were rough for you, well, try dipping into the shoes of the developers who make software based on Facebook and Twitter APIs. Sure, you missed all of Ashton and Demi's tweets for a few hours. But for devs, the two goliath social networking services are their livelihoods. And what's surprising is how Facebook and Twitter left them completely in the dark.

Over at the Facebook Developer Platform forum, app developers began discussing the timeouts and connection errors immediately at 7a.m. Desperate to know if the problem was Facebook's or their own, developers streamed onto a thread about slow-loading pages that shot to the top of the forum's activity list. "Ajax calls are always timing out for me, can anyone else confirm this?" said one dev. "Timeouts and failures everywhere," replied another.


Not until an hour and a half later did anyone from Facebook reply to the thread to confirm the problems: and when they did, they didn't give any answers or information. They simply referred developers to Facebook's tersely, infrequent platform updates, which were more PR-speak than technical support. (Oh, and "sorry for the inconvenience.")


Twitter handled the situation in a more personable manner, but still left its developers largely in the dark about what was going on. That's not only rude, but counterproductive for the platform; as some of Twitter's developers pointed out, having a bunch of apps making repeated, unanswered API calls to the server while repairs were underway only innundating the platform further, making it harder to get things up and running.


Certainly in a time of networking crisis, developer relations is bumped down to second place—the problem at hand commands all hands on deck. But why not spare five minutes to keep developers apprised of the situation? These are the people that make the platform viable; indeed, they're the ones who make it a platform. They're also incredibly talented, caffeine-addicted and innovative, and they want nothing more than to help get things running again. Even if there's no practical way to lasso all that enthusiasm to solve a problem, there's no sense in stunting it just because it might mean stealing five minutes from your engineering solution.

Add New Comment


  • Chris Dannen

    @ R T, that 7:26AM update read:

    "Network issues
    Aug 6, 2009 7:26am
    We are currently working on a network issue which is likely to be the cause of timeouts being reported for Platform. We will update when the issue is resolved."

    I don't see how this is much help to developers.

  • Emeri Gent [Em]

    I fed my comment through the Twitter Tweet machine & so it is churned into 8 additional tweets below :-)

    In regards to this particular outage, I will be surprised if the contingency coverage for developers does not eventually improve.

    The developer ecosystem in this market is as important as the developer ecosystem in any market a.k.a such as automobiles etc.

    The key today is how adaptive we can become for the old days of lumbering tankers is over, today is the age of the speed boat.

    We have to find ways to improve where & why we turn, when we shall maneuver & how we can move through greater productive cooperative speed.

    The key here isn't the speed of response, the reason for the PR speak is fear driven.

    We really need "Lawrence Lessig" kind of thinking about new frameworks that make contingency a cooperation rather than a planning exercise.

    In a rapid change environment, what else causes the rise of fear when we confronted to adapt to change?

    We must learn how to zig when if we learned only to zag...

    And that is what I tweetin-think...