Ryan MacMillan

I’ve been watching the Google Campfire video on their recently announced OpenSocial initiative. And aside from it being wonderfully cheesy - there literally is a campfire and the smoke continues to waft across the presenter throughout the presentation - its is a handy overview of the situation that currently exists with social networks and the different APIs they use and how OpenSocial deals with that.

There is a lovely parallel (in my mind at least) with the Australian rail system and its development in the mid 1800’s. The first rail systems were designed to serve individual cities as independent systems feeding hinterland produce to the coast. The idea that they might connect with other city and therefor state systems never occurred to them. It is interesting to note that it was a combination of a the lack of sufficient independent venture capital and of local authorities working independently - Australia was not a federation of States until 1901 - that resulted in local Governments developing their own systems independently and with the resulting lack of interoperability. By the time the states were joined by Federation three different gauges had been used with the result that in 1917 an east west crossing of the continent required six train changes. Along with the poor user experience there were the other issues of incompatible equipment and operating practices. The legacy; there are still three different gauges in Australia - there’s clearly quite a switching cost - and it was only in 1995 that the system allowed for trains to travel between all of the capital cities. If you’re ‘that’ interested, there’s more here.
The situation we have today is that with the rise of social usage of web, developers are now creating applications to work on these social sites by accessing the APIs of each site. But developing for many social sites means that they have to develop for many APIs each with the own variables and variations - this often leads to decisions to limit the number of sites an application will work on and the users experience is a far from seamless one switching between the various services such as the two I tend to use; Linkedin and Facebook.

Those who refuse to learn from history etc etc…so Google’s point is that by adopting the Open Social standards we will avoid the problems for the Australian rail service; different rolling stock, limited to the systems they were designed for using different operating practices resulting in an overall crap user experience.

The purported benefits are that developers will have maximum distribution for their applications (at the launch of the service the partner sites participating represented 200 Million users - Andreessen from Ning refers to 100M so someone should do the math properly). It also means the web site owners, by adopting the open social standards for HTML and Java (fairly ubiquitous technologies) will be able to run the applications the developers produce to these standards on their sites. And of course consumers then get lots of lovely applications that work seamlessly on lots of sites.

But why do Google give a damn? And this gives you some insight into the scale of their thinking - if the web is a better place - more people will use it. More people - more searches - and other application usage. Simple really. Like car companies and suppliers supporting and part financing the development of the Lincoln Highway in the States. (I have no idea where the travel thing has come from - lets see how long it lasts).

Marc Andressen’s company Ning is a participant in the OpenSocial initiative and is already running OpenSocial versions of applications such as ILike which is then also running on other social networks such as Orkut and Hi5. He explains it quite elegantly on is blog and there’s a explanation of how to add a gadget to a network on Ning.