ReadWriteWeb

R/WW Thanksgiving: Thank You Google for Open Social (Or, Why Open Social Really Matters)

Written by Alex Iskold / November 22, 2007 8:39 AM / 13 Comments

When Google and others ganged up on Facebook a few weeks ago, to many of us, Open Social looked like a marketing move. The news came suspiciously close to Facebook's ad platform announcement and after a close look, the API looked very raw. Most participants just announced their support without having any concrete implementation.

Yet, Open Social is not a fluke and neither it is an accident. It is an important step in the evolution of social and open web, a step that we have seen taken before in other circumstances. It is called commoditization. By creating an exchange of gadgets and social information Google and Co. declared that the era of social silos is over. In this post we look at the details of the open social API, discuss its adoption and look into the future of the social web.

What is Commoditization?

Commoditization is a standard point in any technology cycle. First, when we invent something new it is scarce. Think back to the first cell phones. The prices were high and people were standing in line just to get one. As the technology gets refined and phones become cheaper to make, more of them are thrown into the marketplace. More companies get into the act and prices drop. The end result? Seven year olds are running around with cell phones today because they are free.


The diagram above is from Doc Searls.

Cell phones became a commodity just like TV and transistors and DVD players and even computers. Once something becomes a commodity it turns into a platform on top of which new stuff can be built. Because the majority of the US population has a cell phone, there is a big opportunity for the wireless platform - calling, accessing the web, playing games, and so on. Commodities also drive standards because there is no competitive know-how left. If everyone knows how to make cheap phones there is no reason to tie the phone to a particular service. Granted, in our telecom industry things are much more complicated, but you get the point.

Historical Look at Java-based Application Servers

One of the most recent examples of commoditization in the software industry is the rise of Java-based application servers. An application server is a container for running enterprise applications. It is, in essence, a sandbox that abstracts away enterprise services - application deployment and monitoring, database access, web access, queuing, security, etc. Initially several incompatible products were on the market, most notably from BEA and IBM. The incongruence between vendors meant that companies could not develop portable applications - each application had to run on the software from a specific vendor. This created a lock-in and reluctance in the marketplace.

Things changed when, under the leadership of Sun Microsystems, the Java community authored a single standard specification for an application server. Sun, Oracle, BEA, IBM and many others implemented the specification in their product. This meant that each application server now had a standard application packaging, deployment and execution. Because everyone complied, the applications became portable. Well, almost - the truth is that vendors still have proprietary things in their systems, but it is much easier to move applications around.

What is Open Social?

Simply speaking, Open Social is like a specification for Java Application Servers except that it is a spec for hosting applications and widgets inside social networks. Specifically, Open Social defines the following three broad areas:

  • Widget/Application API
  • Friends API
  • Activity API

The Widget/Application API is similar to the Facebook Application API. It defines a protocol between widgets/applications and the social network (the container). A widget written according to the protocol is guaranteed to work in any social network that supports it. As an example, if you write a widget for iGoogle it will also work on Ning. This is because both of these environments have implemented support for a common widget format. This format itself is very straightforward and is based on the existing Google modules. Each module is an XML file which describes the properties of the widget and contains JavaScript code for it. The deployment process typically consists of putting this file on a network and then specifying a URL on the social network's configuration screen.

The Friends API allows external applications to query a piece of the social graph inside the social network. Based on explicit permission from the user, applications may get a list of his or her friends. This is important, because it unlocks the social graph. For example, Flixster holds a piece of your social graph that defines people who like similar movies, basically your movie friends. If Flixster implements Open Social then you could send this information over to Netflix. This is powerful and important, because it empowers consumers to control their data and saves people time by helping information from one service bootstrap another.

The Activity API is designed to both import and export the user activity from the social network. It is similar to Facebook's news feed and Beacon ads, except that those are not exportable. Examples of this include things like: Alex posted a photo to Flickr or Alex twittered that he is on a plane. And also for the activities inside the network: Alex and Richard just became friends or Alex is attending the NYC Tech Meetup.

The first API is really about portability of widgets and applications, while the second and third are about portability of the social graph and attention information. All of these still have a long way to evolve but they bring powerful concepts of openness and interoperability into the social network marketplace.

What is the Impact on Developers?

The slogan: write once, run anywhere, is music to the ears of any developer. Having to support a piece of code that is targeted towards many incompatible platforms is a nightmare. With Open Social, widget and application developers will be able to have a single, hopefully simple, code base. The direct benefit is that developers can focus on the essence of their offering instead of having to deal with infrastructure and portability issues.

The fact that the same piece of code can run on many social networks is also likely to encourage more companies to enter the widget marketplace. Prior to their announcement of Open Social some companies may have been reluctant because the cost of creating specific widgets for each network might have been prohibitive. With Open Social, there are no longer any excuses. Investing in a widget or a widget set that can be deployed across a wide range of social networks is a no-brainer.

So What About Facebook?

Commoditization of social networks is not good for Facebook. It has been positioning itself as a leader, a unique place that connects people in the just the right way. The Facebook platform is what made Facebook into "the company" of 2007. If everyone has the platform and not a proprietary, but standard platform, then Facebook's value shrinks back to the size of its current audience.

Despite the fact that Open Social is a big stone thrown at Facebook, it is still doing fine. With close to 60 Million users and the deal with Microsoft under its belt Facebook is likely to continue its march forward. It is also likely to join Open Social. Being an odd duck in this situation is not going to fly well with customers. There is a chance that it is going to play the Apple "we are the best and closed" card, but it is a rather small one. Facebook is not going to loose much from joining Open Social. In fact, like IBM some years ago embraced the open source, Facebook can do the same with Open Social. If it does, it might actually own it and cancel out the impact that Google and Co. is trying to make.

Where Do We Go From Here?

Open Social is a major step towards opening up web silos. As we have previously discussed here, most companies on the web have been silos. From Amazon to Netflix the companies have held on to the user information that they gathered, turning it into a business advantage. Open Social paves a way to a potentially new kind of web culture. In that culture, companies would recognize that users are entitled to their information. It should be importable and exportable. It should not be locked in.

If Open Social is implemented on the broader scale, there is likely to be a cultural shift. Consumers are going to recognize that if their social graph is portable and if their attention information is portable in social networks, then it should be portable at large. People are going to demand that their Amazon purchasing history and Netflix rental history is accessible via open API. If that happens, we will effectively enter the age of the attention economy.

Now tell us what you think of Open Social and its implications. What companies would you like to see join this initiative? What do you think about Facebook's play, will it join the initiative or remain closed?



1 TrackBacks

TrackBack URL for this entry: http://www.readwriteweb.com/cgi-bin/mt/mt-tb.cgi/1785

Comments

Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts

  1. Form what I have read, the API is still one way, you cannot write data, and the data cannot flow between networks. There has been no offical call from google or any other participants that it will happen in future. So how does it fit into what you have just said? The information is still locked in silos? OpenSocial cannot free the data, we need new and better ways of dealing with identity itself, like the openID initiative.

    Posted by: Elf | November 22, 2007 9:22 AM



  2. Agree with #1.

    In course of our deploying our new site (China-centric),
    we encounter the same observation, and we've resorted to
    supporting OpenID *in addition* to OpenSocial... fyi, /ac.

    Posted by: ac@113.com | November 22, 2007 9:36 AM



  3. I'd love to see everyone embrace Open Social, and most importantly, much more, Brad's project, I didn't read the details, but his project could give users an overview of all his connections, manage data access permissions, etc.

    Tired of millions of social networks and data duplication, all your photos, videos, etc should be treated as unique assets, associated with you. Social networks would provide storage for those assets, and their own unique way of creating relations between assets, and between people.

    Also, Brad's project should be used to manage more than social data, ALL of one's data should be managed though there, relations between banks and you, driving license, etc etc, they're all assets you'd own/be associated with, grouped in macro channels, each corresponding to a kind of asset, like, the bank channel, you'd query it to get a list of associated banks(granted you had permission to), comms channel, etc, physical data ownership channel, I think you can already get the idea and imagine some convenient situations where this kind of central database would be handy, and please, don't be paranoid about privacy( like everyone else will be if such thing is ever created).

    Just my opinion, sorry if it's unprofessional or something like that, guess it's more like an end-user p.o.v.

    Posted by: lmjabreu | November 22, 2007 9:38 AM



  4. We cannot wait to be included in open social!

    Posted by: lapp | November 22, 2007 9:56 AM



  5. @1 and @2

    Here is the references to the APIs:
    http://code.google.com/apis/opensocial/docs/gdata/people/developers_guide_protocol.html

    http://code.google.com/apis/opensocial/docs/gdata/activities/developers_guide_protocol.html

    Posted by: Alex Iskold | November 22, 2007 11:56 AM



  6. I think you're right; Facebook will join OpenSocial (when they are good and ready!)

    Posted by: Mark | November 22, 2007 1:42 PM



  7. As far as I know OpenSocial hasn't freed the social graph and has no imminent plans to do so. Rumblings elsewhere however suggest that this is on the cards.

    Facebook should definitely join. Then it can focus on being the best social networking site instead of the only one.

    Posted by: Charlie | November 22, 2007 1:57 PM



  8. #1 and #6, at first I agreed with you and then I checked out the docs again: http://code.google.com/apis/opensocial/docs/gdata/people/developers_guide_protocol.html

    People Data API. It isn't released yet but some people have already implemented bits of it; http://www.voidstar.com/node.php?id=3044

    Very rough, lots of problems (security is non-existent) but there you go, a free and open graph.

    I'd love to see MySpace and Linked In implementing it though. They'd be sucked dry in minutes by ever bot on the net.

    Posted by: Paul M. Watson | November 22, 2007 3:52 PM



  9. A couple of weeks ago i've join to facebook for the first time .. It's really really cool.. i'm from korea. we usually use 싸이 like facebook site for social network. I've exprienced really fresh someting like this. And as your opinion i wanna also facebook to be open social.
    Good topic.
    Thanks.

    Posted by: Frank | November 22, 2007 3:57 PM



  10. Googe effort seems like IBM effort on ODF, trying to playing politics with open card where they are defeated.

    Google should do open search, this will help the world faster. Why wasted resources on some thing that you are losing. Just accept that you are defeated by FaceBook. Blogger is losing to Wordpress, now Okurt is losing to FaceBook. Big but weak.

    Posted by: TanNg | November 22, 2007 9:57 PM



  11. This might sound naive but OpenSocial sounds more like a guild than a user enhancement initiative. It's fine to get on the same level and play the game but when have corporations agreed to create environments that will help each other.

    Posted by: mithun | November 23, 2007 1:26 AM



  12. As soon as Google announced Opensocial API, NYT published an article named "OpenSocial is not a Facebook Killer". And it's not. This explanation remembers it to us. Thx!

    Posted by: Mariana Oliveira | November 23, 2007 3:50 AM



  13. You state:

    "People are going to demand that their Amazon purchasing history and Netflix rental history is accessible via open API."

    I really doubt it. Other than a tiny subset of companies or developers or bloggers, what consumers in the real world have ever asked for this? As long as their favorite sites work they don't need or want their data shared with others.

    Also, what successful company in a dominant position has ever voluntarily released their core assets (aka crown jewels) to an open standard? Answer: None.

    In your world, here are some other things that would be great if they were open:

    1. iTunes supports the Zune and other non-apple music players

    2. Google opens up AdSense so Microsoft and Yahoo can serve better ads for Google's 1+ million member content network

    3. NetFlix allows WalMart, Amazon and Blockbuster to share the member data so consumers enjoy more convenient movie rental and purchasing

    The only way a for-profit company would do this is if the government makes them. The one example I can think of is when Xerox was forced to give up all their patent rights in the copier space to a host of Japanese competitors. This was great for lowering the price of copiers but not so hot for Xerox. One footnote, is that this forced disclosure of their copier patents in the 1970's was the reason that Xerox undervalued patenting or even agressively copyrighting their inventions in the area of graphical user interfaces with a desktop metaphor, document windows, dialog boxes, icons, mouse, pop-up menus, etc. Xerox shareholders missed out on what could have been some of the most lucrative software licenses ever because their management learned the wrong lesson about protecting corporate IP.

    I doubt Facebook will make the same mistake.

    Posted by: Lee Lorenzen | November 23, 2007 8:48 AM



RWW SPONSORS


FOLLOW @RWW ON TWITTER

ReadWriteWeb on Facebook



TEXT LINK ADS