In part one, Hackman talked about Tibco's vision of SOA business applications and its Ajax Message Service (AMS), which is designed to provide event-driven SOA by pushing live data and events from servers to Web pages, Ajax applications and other software.
How easy has it been to tie what you've done in General Interface into the larger SOA message that Tibco has?The whole impetus behind Tibco's acquisition of General Interface in 2004 was that relationship. General Interface was one of the first to show SOAP communications from JavaScript in the browser back in 2002. It's really been a pretty seamless integration. People tend to think of SOA as an integration strategy, but are realizing today through the advent of Ajax that it is more of a generalized architecture approach to solution development.
If you look at the harbinger of all this Ajax stuff, Google Maps, it's not just a JavaScript client that's phenomenal, it's the service behind that. You've got a really rich set of data points that you can query against with a simple API and bring all that information back. But it's the symbiosis between the service behind it and all the rich Ajax client work together. The difference though is that Google Maps is publicly accessible. They don't have to deal with all the security issues, and it's one service. When you go into the enterprise
Requires Free Membership to View
When you register, you'll begin receiving targeted emails from my team of award-winning writers. Our goal is to keep you informed on recent service-oriented architecture (SOA) and SOA-related topics such as integration, governance, Web services, Cloud and more.
Hannah Smalltree, Editorial Director
what happens when you have five or 50 or 100 services. If you open up SAP you've got 5,000 services. Those are the kind of enterprise scale issues that Tibco Active Matrix (Service Grid) deals with, so that all the policy and governance and the orchestration of those services with that number and complexity and two decades of legacy, the ability to expose that up to a human being so they become part of the flow of information in an SOA is exactly what General Interface is about. Do you think people doing SOA believe you have to have an API? Do you think we need to start thinking a little broader, perhaps? People are thinking at a variety of levels. We have customers who use SOAP in 50 percent of their efforts. Some people automatically associate Web services with SOAP, but they do that when there's anticipated a high degree of reuse for that effort. They get the return on that effort from taking the extra time to create that, but some organizations are deploying Plain Old XML for something tactical or solutions where they know the services and solutions aren't going to have a high degree of reuse, therefore they don't invest in that overhead. Both are example of service orientation, decoupling the client, no longer tightly coupling the HTML page to the application information. But whether they go that extra level to embrace the full WS-* standards or UDDI is based on the application and the intent of the application. So we see both. Large caps SOA and lower case soa implemented in organizations. Are you running into REST these days?
Absolutely. Again in terms of simply implemented types of services without any overhead of SOAP, people use REST. People are starting to use more JSON. We've sponsored the development of DWR, the popular Direct Web Remoting library for Java to JavaScript remoting. It's really an instance of JavaScript services. We really think with all these things what's common about then is they're accessed over HTTP. That's the great normalizer. There's a variety of techniques that people still want to use. What do you expect will be coming out of the Open Ajax Alliance in the near future?
Tibco is one of the steering committee members of the Open Ajax Alliance. I participate heavily in that organization. The key project right now at the Open Ajax Alliance is focused on Ajax interoperability, enabling the whole mashup phenomenon that's taking place with multiple widgets from multiple toolkits coexisting in the page and playing nice. It turns out that when you go to script these widgets together you have a classic server-side integration problem – that n-squared mathematics you run into. The solution to that problem turns out to be a service bus. So it's really interesting that the Open Ajax Alliance for facilitating inter-operation of these mashups has also arrived at publish-subscribe architecture. There's a project called the Open Ajax Hub, a publish-subscribe event hub – what we did was contribute the core of Tibco's page bus to implement broader messaging, subscribe bus functions. So there's a very robust infrastructure, all of it under three kilobytes of JavaScript. It's tiny, but extremely powerful and that's what publish-subscribe is supposed to be to facilitate the interaction of Ajax components and libraries.
 |
||||
|
|
|||
|
||||
That's the question. Obviously, time will tell. The types of problems that seeks to solve tend to be more this whole sort of cinematic user experience that's out there that Flash has addressed in the past or what (Microsoft) SilverLight is trying for. It's pretty transparent that they are trying to displace Flash. There's a little new plug-in war, if you will, that's going on in those areas.
The core thing about Ajax is it works with the technology you already have. The adoption rate for Windows XP was eight years to get above 60 percent. The nice thing about Ajax is it works on Windows 2000 workstation server. So there's a real value to working with old tech. What's cool about Ajax is it's all these brand new capabilities using these old technologies. In fact, that's exactly how the Wall Street Journal when they covered Ajax for the first time in mid-March of 2005 described the Ajax phenomenon. Ajax has a potential very long life because you can use it today rather than have to wait for tomorrow's adoption.