« TalkR | Main | An AJAX framework... »
20060629 Thursday June 29, 2006

Integrating with Industry

I've talked here about the group I work in at Sun and how we build common components and services. Most of what we do allows external customers to integrate with Sun's services to manage their hardware and software that they get from Sun. All of those things require the user to authenticate using a Sun Online Account. Sun has a long history with our online accounts and the systems that have handled this type of authentication. Needless to say there has been a lot of evolution in this space. The most recent of which is attempting to expose public webservices to get at the functionality. This latest attempt is something that follows the same history of identity systems at Sun, it is a completely proprietary implementation of webservices around identity creation, management, and validation that makes interoperability somewhat painful. This does use the latest version of Sun's Identity Management suite but does not expose any of the Identity Federation standards available there.

Since I've had to deal with integrating with many of these back end identity systems at Sun in my time, it always interests me to see what the rest of the industry is doing in this space. So when a friend of mine sent me a link to this ZDNet article I found it pretty interesting. Essentially Google is creating their own pseudo identity federation implementation and opening up the api's to get at it for anybody to use. Now Sun has their own set of users, but what if we were to say that we were going to just start deploying all of our services as Google services and just start using this new authentication mechanism. We could get completely out of the business of managing customer identites as well as potentially a lot of other things. We could even leverage the infrastructure that Google is building out (read GooglePlex) and save a ton of money on data center space by delivering our services through them.

Before you rip this apart with all the potential problems, like alienating other vendors (Yahoo, etc), privacy issues, delivery issues, quality of service, etc, etc, etc..... just know that I realize that. I'm trying to make a point, which runs through a lot of my blog entries, that inter company integration can greatly benefit a lot of things. Why don't we let the people who do things the best do them and quit trying to do a half ass job at them ourselves? Also, some of these companies are doing some really cool things, and they have gone out of their way to make it blazingly easy to integrate with what they are doing, why? because it benefits them too. Take a look at these developer pages at Amazon, Google, Salesforce.com, Intuit, Yahoo, and I'm sure tons more. If we wrote service applications to service Sun's hardware and software and deployed them through integrations like these, there seems like there is so much potential. Probably everything we would need isn't available, so we build the missing pieces. But the value and potential of integrating with these things is endless, and the cross sell opportunities are endless as well.

So why don't we quit duplicating effort? Let's quit following the industry and figure out where we can best lead and where we have the best opportunity to innovate.

Bibliography:

Posted by mikedav ( Jun 29 2006, 09:57:06 AM PDT ) [Listen] Permalink

Comments:

Post a Comment:

Comments are closed for this entry.