On OpenSocial

It’s likely this is the umpteenth article you’ve read about Google’s OpenSocial initiative, so I’m going to see if I can’t make this one worth your while.

First – what is OpenSocial? Briefly, OpenSocial is a set of Google-defined API’s and methods for passing “social” information between two applications. Therefore, one application is the “datastore” and another is the “view.” These applications interact through (and are limited by) methods defined in the OpenSocial API. Notably, these API’s are platform agnostic – while they are built on Google’s datastore, implementers are not limited to Google’s datastore.

Why does it matter? The good, non-fluffy side of Web 2.0 is fueled by mashups and API’s. The problem with API’s is that they’re inconsistent, write once, etc. Google is attempting to end-around one-off social API’s by creating their own, with the false guise of standardization.

But wait, this can’t be the first “open” Social API, can it? No – there have been a number of attempts at creating standardized wrappers for social. FOAF is a notable example, XFN another. Like Google’s initiative, none of these are “standards” in the IETF or W3C sense. Seizing on its place in the market, its powerful, hegemonic voice and the cadre of willing followers, Google’s opensocial is a super-standard – not sanctified by any body other than the sheer volume of developers who will start playing with it.

So isn’t Google just creating a new set of problems? Yes and no. Social information transfer should be standardized, but the standard bodies (and interested parties) have simply been too slow to create and approve standards. It’s the end of 2007 – we’ve been using SNS for years now, so there’s no excuse. Of course, these social systems need a standard and mode of integration to be useful, and Google is providing both. This is a path towards standardization, and it would be useful if Google did carry through on a community-supported standard. The standardized passing of social information is extremely valuable, and third-parties need the support of standards bodies to justify investment of significant resources. Until then, everyone is at Google’s whim, which is a tenuous business proposition.

So what is Google really trying to do? By placing “opensocial” in the open, Google is demystifying how it will interconnect its properties. This is as important strategic move; Google contains so much personal information about all of us that openness will benefit the company when Google decides to interconnect. First, Google will be a follower – the interconnections will be old news, and precedented throughout the web. Second, bloggers and information purveyors will be able to make Google’s case – look at the effort I’m taking to explain OpenSocial right now. When Google interconnects I’ll just point to this article and tell you it’s been coming for months or years now. The techno-elite may be ok with having the walls broken down between Google’s services, but on the open web, it’s going to be traumatic. Google is preparing for this with OpenSocial.

It’s larger than platform interconnectivity and framing, however. By getting developer buy-in and deploying these “standards”, Google’s also developing bankable data sharing architecture between countless third-parties. But Google doesn’t need to be in the middle, you say. Of course not…but knowing Google, they will find a way to incentivize buy in, with some really cool innovation. I can already see the OpenSocial maps parser or something like that. Google will find ways to make themselves centric in this scheme – they always do. And that’s not evil – that’s just a corporation making a strategic business move to collect its primary resource: your data.

I concede that opensocial is sorely needed and is a long time coming. We need better ways of interconnecting – and good, earnest people have been working on this problem for years. Closed platforms like Facebook’s are very Web 2.0, but global interconnection is whatever is after Web 2.0. It’s unfortunate that Google, rather than a standards body, is at the center, but perhaps this will outline a standard way forward. Make no mistake – it is Google’s powerful voice that makes them capable of declaring such “standards”, but OpenSocial is really just Google’s take on the matter. Of course, Google is more equal than others, so this is far from the last you’ll hear on the matter.

Update: I like Dave Weiner’s take. “Standards devised by one tech company whose main purpose is to undermine another tech company, usually don’t work. In this case it’s Google trying to undermine Facebook. And I don’t think it’s going to work.” Well put.

Tags: ,

5 comments

  1. Fred, maybe you’re the right person to answer this question I asked on o’reilly radar:

    “Is this just another widget platform? Or is it the first step towards the open social graph? I guess I’d assumed from the pre-launch hype that it was the latter, but all I’m really seeing right now is the former. Am I missing something?”

    [OT: I find it ironic that your blog provider does not let me sign in with openid ;)

  2. It is a platform – but it’s no more a step towards open social graph than XFN or FOAF. So I’d agree with you. This is primarily Google’s attempt to claim real estate in the social space – something they’ve done very poorly.

    And yes…pretty much everything is ironic about my blog provider..but I was using them before they were owned by Google. One of these days I’m going to migrate to my own domain…

  3. Luis, I don’t think you’re missing anything. While Google may yet make strides towards the “open social graph,” this isn’t anything of the kind. It’ll be nice to develop one app (widget if you prefer) that can run on a number of sites, and I’m intrigued about how LinkedIn will be using this, but it’s still just another widget platform.

    Also, does this mean Google will be allowing open social apps on my Google homepage?

  4. “So what is Google really trying to do? …Google is demystifying how it will interconnect its properties.”

    That’s secondary.

    First, Google wants to create a marketplace and enabler for Social Adsense.

  5. Hashim, I agree. Here’s a comment on left on Scott Kevton’s blog:

    Scott, I think you’re right about the Google’s centrality in the mix. More than anything, this is a darkweb data-mining effort; they’re going to benefit from all of the also-ran social nets that will expose user data in the name of “openness” but are really hoping to use this as a strategic in to Facebook’s market. Unfortunately for them, this won’t afford competition with Facebook, but Google will get all of the data.

    As Google’s algos are already graph based, they’ll be able to quickly manage this social graph data. In fact, the Spertus et. al. publication from 2005 Evaluating similarity measures: a large-scale study in the orkut social network pretty much lays out Google’s strategy. Just like Google Analytics, the DoubleClick acquisition, Adsense cookies, the Google toolbar – it’s all about collecting as much of our data as possible. This is a natural area for expansion of their data collection programs.

    Where we differ is our trust level of Google. Google may uphold good data maintenance standards, but they’re a public corporation who is only responsible to the bottom line. I worry about what might happen to our data one day.

Leave a comment