In Defense of Walled Gardens

It’s easy to assume that when it comes to data and software development, “open” is always better than “closed.” We’ve seen an explosion of open source software, praised companies for supporting open standards, and breathlessly tracked products with “open” in their name, from OpenID to OpenSocial. “Closed” has become the scarlet letter of the Internet, at times expressed by the censure of being branded a “walled garden.”

Facebook has often faced this criticism, particularly after unveiling the Facebook Platform in 2007. Several bloggers compared Facebook unfavorably to AOL of yesteryear, eschewing Facebook’s “proprietary” (gasp!) FBML and FQL interfaces. Some even portrayed Facebook as a competitor to the Web itself. While the definition of “walled garden” was not always particularly clear, observers were unhappy with so much data flowing into Facebook and so little flowing out.

One would think that now, with the Facebook API able to expose your wall, News Feed, inbox, and just about every bit of profile data (even e-mail addresses to some degree) Facebook would be allowed in the open club. Indeed, some writers have noted changes since 2007 that justify dropping the dreaded horticultural moniker. But others continue to speak worriedly of Facebook’s dominance, even still drawing comparisons to AOL.

I, for one, not only have full confidence in the Web outlasting any supposed competition but also see Facebook as very much a part of that resilient network. In fact, I’d like to propose a bit of Internet heresy by according walled gardens a place among the open fields of the online realm.

First, let’s establish one fact: Facebook has always been part of, not opposed to, the Web. Disregard ridiculous arguments over FBML and FQL, which were no more of a threat to HTML and SQL than Smarty and WordPress template functions. Open any Facebook application, choose to “view source,” and all you’ll see is good ol’ HTML, CSS, and JavaScript. The Facebook Platform allowed developers to build on top of Facebook, just as Movable Type and Joomla allowed developers to write plug-ins using Perl and PHP. (Differences: you could not roll your own Facebook, Facebook essentially installed every plug-in, and you have to host the code.) That Facebook disallowed certain HTML security risks and added a few convenient tags for interfacing with their content in one approach to development (one could always write full-blown HTML using canvas iframes instead of FBML) hardly meant they were reinventing Web standards.

Technical considerations aside, some writers argued that Facebook opposed the Web in spirit – more specifically, the spirit of openness. Even though Facebook applications (and inverting Facebook’s criticized original setup, other web sites via Facebook Connect) have wide access to Facebook data now, average users still face hurdles if they wish to view posts and information from other Facebook users. At minimum, one has to create an account and login to Facebook to see beyond bare basics. Prior to recent privacy changes, access to content from non-friends was highly limited even after logging in. And while some of Facebook’s data should start appearing in search engines this year, very little has been indexed by Google so far. As I said before, users generate much content within the context of Facebook, but that context usually remains locked away from public access.

Before I respond directly to such charges, I’d note that Facebook (or any social networking utility) serves a limited purpose. Did you catch that? Facebook serves a limited purpose. Facebook was never meant to duplicate the Internet. If I need reference material on world history, I might turn to Google or (as a starting point) Wikipedia. If I want to know the latest technology news, I can bring up Techmeme. If I want to catch up on a favorite TV show, I’ll probably load Hulu. None of these tasks have any inherent social component that would cause me to first open Facebook when fulfilling them.

Last summer, however, I wrote a series of articles on particular doctrinal issues that affected certain people in churches and organizations I’ve been a part of. I’m not ashamed of my opinions, but they involve points that would not make sense to someone who did not have the background and context of the limited audience I had in mind when writing. Consequently, I would prefer such musings did not appear in the Google search results of an acquaintance unfamiliar with my topics. I shared my thoughts with certain friends via Facebook Notes.

I have friends who live several states away that wish to keep me and others posted on life in their growing family. They want to share what adventures their children are having with extended family across the country. Friends desire to see how they’ve decorated their new home and exchange tips on managing it. Rather than open themselves to potential hazards of their house and kids being featured in an image search, my friends can use Facebook’s photo albums to control who can observe their daily life.

These are but two use cases out of a hundred or more that (1) inherently involve a user’s social graph and (2) inherently involve content not intended for public consumption. To argue that Web-based services other than Facebook could provide similar functionality in an open context misses the point. Yes, I could have published my articles with Blogger, my friends could post their photos on Flickr. But these particular examples are not simply about sharing ideas and pictures – they involve sharing ideas and pictures with certain people.

Are you dissatisfied with Facebook limiting access to content? That’s where the rest of the Web (again, Facebook is one part of the Web) comes in handy. If you’re a blogger who wants the world to hear your thoughts, forget Facebook and start a blog. If you’re a photographer who wants to advertise your portfolio, forget Facebook and use a more open service. If you’re looking to interact with a small subset of the world, however, a walled garden may be just the thing for you.

Of course, these days Facebook’s leadership may cringe at my last paragraph, as they seem to be taking a new angle on their service’s purpose. But a few years ago, privacy and control (in essence, the very things that made it a “walled garden”) are what distinguished Facebook from competitors. Personally, I have trouble buying Mark Zuckerberg’s story that “if he were to create Facebook again today, user information would by default be public” (ReadWriteWeb), as such a site really wouldn’t have been much on an innovation. Recall that prior to Facebook’s rise, MySpace dominated social networking sites, and many (if not most) MySpace profiles were publicly accessible. Limitations are what made Facebook novel – originally, only college students were even allowed to create profiles on the site, and all profiles followed a strict layout.

In my experience, friends flocked to Facebook because it let them participate in new technologies (sharing digital photos, for instance) but in a controlled environment where they could enjoy a level of privacy. The garden walls were selling points – college students didn’t want just anyone seeing their photos and messages; later on, parents didn’t want their teenagers communicating with just anyone. I think that’s partly why Facebook’s recent moves encouraging users to share more openly generated such controversy. Users felt they had fallen victim to a “bait and switch” scheme: they invited their friends to use Facebook so they could share privately, now suddenly Facebook has forced them to share certain information and is pushing them to share the rest.

It’s also worth noting that in its early days, Facebook offered little functionality that couldn’t be found elsewhere. The notion of a profile, the ability to send private messages, the exchange of ideas centered around certain topics – all of these features were hallmarks of forum sites for years. But while the members of a forum formed a particular social graph centered around a certain niche community, Zuckerberg portrayed a user’s social graph on Facebook as a mirror of their everyday, real-life connections. Facebook (and other sites, such as MySpace) took the features of forums and adapted them for a general purpose audience, where you essentially chose the members of your forum based on who you already communicated with offline.

This brings us back to Facebook’s limited purpose and why some of the hand-wringing over its “proprietary” nature strikes me as overreacting. For instance, back in 2007, Gervase Markham of the Mozilla Foundation expressed concern that the messaging system in Facebook or LinkedIn might turn e-mail into a closed system incompatible with outside domains – in short, a walled garden. But Facebook messages could never replace SMTP e-mail. If I want to interact with close friends, Facebook messages provide a convenient, hassle-free means of doing so. Yet if I need to exchange notes or documents with acquaintances, large groups, or even businesses, Facebook messages are hardly up to the task. Once again, such use cases do not involve my social graph. I think Facebook recognized this as they expanded and started trying to juggle multiple social graphs beyond a user’s closer friends, since messages from fan pages (essentially, communications for a business) are filed away in a folder quite separate from a user’s main inbox of messages.

All this being said, please don’t think I’m opposed to more “open” approaches to handling my social graph, such as distributed social networking – far from it. I think Facebook is still an early player in online social networking, and that we’ll see many more platforms and ideas develop in years to come. But I think we’re still a long way from a time where the open alternatives provide end users with more value than walled gardens in the types of use cases I’ve already outlined. As much as I’d like to see federated social networking platforms thrive, I foresee many hurdles that have yet to be overcome. Distributed networks will have to deal with issues relating to performance (imagine generating a news feed when your friends’ data comes from hundreds of different servers), retention (is data cached, how long, etc.), reliability (what happens when a few of your friends’ servers are down?), privacy (how will access be controlled and monitored), and security (avoiding injection attacks, ensuring all hosts stay up-to-date, etc.), not to mention monetization (a problem that still plagues closed systems). And when it comes to user value, remember that walled gardens have a few inherent advantages – in a security example, if Facebook detects a worm spreading malicious links via messages, they can block all messages with a certain signature or strip out links to a known rogue site.

I suppose my main point is that we need not be concerned if Internet users (even 350 million of them) find use for a service that strikes many technology-minded people as a walled garden. While the Internet was built on open, equal access, that very setup enables some services to provide certain features in a more limited context while still taking advantage of Web technologies. And for many people, these gated communities provide real value that would actually diminish if Google began indexing it all. While certain circles seem to think any notion of online privacy is at best naïve (and granted, some users need to exercise more caution in what they post online, regardless of what service they use), I tend to think that the only people saying privacy is dead are those named in its will. And when privacy does become a factor in sharing online, at times, a garden might need walls.

P.S.: Lest you think I’ve changed my opinion in light of recent privacy controversies, I’d note that I stated very similar thoughts back in 2007 when some of these debates over Facebook first developed.