StrikeIron Blog

StrikeIron an Extended UDDI?

I often get asked if StrikeIron is kind of like a UDDI registry.

Not really.

Conceptually, the idea is similar to StrikeIron's Web Services Marketplace in terms of a location on the Internet where Web services are aggregated and can then be more easily discovered and possibly even bound to by those who have the need to put those services to use. However, UDDI has pretty much been unsuccessful in terms of becoming a public brokerage where people post and find services (Microsoft, SAP, and IBM have shut down their public UDDI registries), and largely has been used internally as enterprise service registries maintaining service metadata.

So why has UDDI seen success internally but not externally? The answer is that there exists a tighter control of standards, behavior, interface, and usage patterns that can exist internally but are very difficult to enforce with the public at large. Left to their own devices, the world will generate metadata mayhem.

For example, if there are ten different services available on a public UDDI registry, they likely each have different sets of response codes (such as "data not found"), different data structures (as simple as the way an address is defined), diverse behavior, inconsistent authentication mechanisms, and varying general usage models.

That pretty much means ten separate integrations must occur, each significantly different. In other words, say goodbye to your old problems (platform dependence) and hello to your new ones (heterogeneous mix of differently behaving services).

However when using internal Web services, this type of control and consistency is much more likely to be in place with consistent metadata across parts of the organization. This means a lot less integration friction and hence a key reason for success internally for UDDI, and failure externally. Internally, a much simpler problem is being solved.

Also, not to mention that with a public UDDI there is no real vetting process of service inclusion, so all kinds of garbage would typically show up in these registries further rendering them unusable.

So what StrikeIron has done is taken those same concepts of consistency and control which have worked internally for organizations and have deployed them in a global public model, or a Global SOA of sorts where interfaces are the same, behavior is consistent from service to service, the response mechanisms are the same, and the business models and patterns behind usage are the same. All of these consistency requirements are strictly enforced, substantially reducing adoption and usage friction.

So rather than doing ten integrations for ten different services, only one repeatable integration needs to be performed, and a much simpler one at that. This is a key component of what StrikeIron offers - plug and play Web services integration with global metadata enforcement. And of course it is combined with the notion that services already exist and don't have to be built like must be done in internal scenarios, a costly and time-consuming process. The services are already there just waiting to be used.

So who says you can't police the world?

Flex vs AJAX

Something that is becoming clear as the Rich Internet Application (RIA) landscape evolves is that there are some religious wars ahead. One that will be particularly interesting to watch is the Flex vs AJAX race as platforms for building RIAs (not that they are mutually exclusive). And as one might imagine, Microsoft is also competing here with its Windows Presentation Foundation (WPF) as part of Windows Vista, but Flex and AJAX are the two frontrunners and the basis of most of the street talk.

What's interesting is that AJAX and Flex are not even the same thing. AJAX is a programming technique using multiple technologies such as JavaScript, XMLHttpRequest, XML or JSON. Flex is actually a product sold by Adobe to help people build Web applications on top of Adobe's Flash Player, which purports a smaller footprint than Java Virtual Machines (JVMs). With its roots in animation, it is the key technology behind the success of YouTube. However, the end goal with each is still the same: rich, user-empowered applications distributed over the Web with the idea of making the Web more interactive.

AJAX has achieved early success because of its utilization of JavaScript which enjoys a large army of proponents, and a great, well-known implementation success in Google Maps for all to point to. Also, its supporters will tell you that AJAX is open and free, and not tied to any one specific vendor (not sure if these are pros or cons). The main challenges that AJAX has is that there are no real standards and it will continue to be plagued by browser incompatibilities and the associated complexity (case in point, the recent update of Mozilla's Firefox broke many existing AJAX applications). And of course there are well known issues such as the browser back button problem with its various ugly solutions, the inability to bookmark pages, and the difficulty of search engine attractability. Finally, how will marketing beat its chest as "page views" become an obsolete metric?

Flex is a solid alternative. The good news is that Flex 2 provides a fairly rich IDE and there exists a good model of Flash Player distribution with its smaller footprint (versus the JVM mess). With a model similar to Acrobat Reader, a developer can often be helped by a vendor providing a stable player. Of course, because Flex is a product, the bad news is you do have to pay if you want any of the tools like the IDE. I believe currently that sells for $500.

Of course at StrikeIron, we are excited about both AJAX and Flex as well as other Web development technologies such as Microsoft's WPF because they all drive the consumption of Web services and create a need for frictionless, easy, and consistent access to many data sources and business functions available over the Web. That's good for us because that's what we do, so for the most part we are agnostic.

But here's one data point. For a couple of my fun hobby Websites, I found myself gravitating towards Flex. Why? Everything I needed was in one place and with AJAX I wasn't quite sure where to start. And I don't quite have the time to test multiple browsers each with multiple versions and updates. For me, spending a little money was worth saving the time. Hmmm....

Bob Brauer
President and CEO

It's Still the Early Dawn of the Data-Driven Web

For quite some time now, StrikeIron has been suggesting that the as the Internet evolves, it will become more and more a medium for transferring data between smart applications, Web sites, and a myriad of new devices, rather than the world's largest virtual tradeshow where everyone has a booth full of product literature.

We are still in the very early days of this gradual metamorphosis, but in the past twelve months we have really seen some exciting innovation that continues to move us all in this direction.

Building smart clients to consume and utilize the seas of data available on the Web, as well as data available internally, has come under many names. These include "mashups", "situational applications", "rich Internet applications" and "composite applications" to name a few, and I'm sure we will continue to see some creative new ones. The common thread between all of the endeavors that the vendors in this space are undertaking is the ability for significantly larger groups of people to build these applications than have traditionally been able to build useful software (meaning business users and your garden-variety Web surfer, rather than just overworked programmers with long work queues already in place), therefore proliferating the utility of the Web and unleashing a powerful new wave of creativity with hundreds of millions of contributors worldwide.

The great news is that the tools available to build these applications are increasing in number, functionality, and sophistication, pointing to an exciting future for this grand amalgamation of the Web.

For example, Andrew Trice of Cynergy Systems built a really nice mashup (Adobe's Flex2) utilizing census data and zip code information data (compiled, aggregated, and made freely available via StrikeIron's Super Data Pack set of data sources), and combining it with Yahoo Map Data. See Cynergy Blog. And here is the application.

Adobe's Flex 2 is just one example of the many tools and groups of technologies giving birth to these burgeoning and innovative forms of reuse. Others enabling the mixing and matching of data sources and functionality across the Web include our own SOA Express (enables mashups to be built in Microsoft Excel), IBM's QEDWiki and Adieu, AJAX (with companies like Zimbra and JackBe leading the way), Kapow's Mashup Server, NexaWeb, Ratchetsoft, Above All Software, and many more on the way. All of these technologies and vendors are giving momentum to the wave.

So the sun is definitely on the horizon, and we at StrikeIron intend to do our best to ensure that all of these platforms, tools, and development environments have consistent, frictionless access to all of the data sources they need to make these applications fly. If we don't have one you need, let us know, and we will do our best to get it.

Bob Brauer
President and CEO, StrikeIron