The Reinvention of Enterprise Software • Java Market Fragmentation • OSGi Bundles • Microkernel-Based Architectures
Eric Newcomer
CTO
Iona Technologies

Eric Newcomer leads IONA's participation in all standardization activities, and has been involved in Web services standardization activities from the beginning. As chief technology officer he is responsible for directing and communicating IONA's technology roadmap, as well as its product strategy as it relates to standards adoption, architecture, and product design.

1.  The large Internet businesses will become a big inspiration for enterprise software innovation as software vendors start to develop products based on requirements from Google, eBay, Amazon, PayPal, etc. This trend will result in the reinvention of enterprise software, and during 2008 this trend will become clear. The traditional middleware products such as J2EE application servers and relational database management systems were developed to handle the load of any business. But the Internet load is much greater than this, as are the requirements to be always available. The old centralized, mainframe-based software designs are being rethought, using the cheapest computers and disks possible to achieve the highest possible levels of scalability, performance, and reliability. During 2008 what is happening here will become generally understood.

2.  In the Java market fragmentation will increase rather than lessen. The recent split between JBI and SCA, and the disagreements over Java EE 6 and OSGi will escalate tensions as the pressure increases on BEA/Oracle, IBM, and Sun to take market share from each other in a diminishing market. Meanwhile, Microsoft has an opportunity to grow stronger behind the leadership of Ray Ozzie and is likely to surprise those who believe the battle for the enterprise is over and Java has already won.

3.  Specifications and reference implementations for the enterprise edition of OSGi software will be completed, laying the foundation for the most significant change in the Java market since the emergence of the Spring Framework, although Sun is likely to continue to oppose it. J2EE application servers will finally become more modularized (buy only what you need) and Java developers will be able to think about enterprise applications in terms of a combination of OSGi bundles, some developed by the user and others supplied by vendors - all of which work seamlessly together and support dynamic deployment and update capabilities.

4.  Resource-oriented computing, aka REST, will finally start to gain serious traction (see also No. 1), although its rabid adherents won't be satisfied with what will be less than total domination (yes Virginia, people will still be using Web services, too). Vendor and user support is on the rise, and more and more people will understand how to take advantage of this powerful architecture. Enterprise applications will start to include both service-oriented and resource-oriented capabilities. OSGi-based infrastructures (see also No. 3) will help the Java world combine both sensibly.

5.  Microkernel-based architectures and lightweight containers will grow in popularity as people gain experience with SOA-based project design, development, and deployment and understand the benefits of "just the right amount of software for an SOA." SOA deployment strategies based on grid and virtualization technologies will also become widely adopted, since lightweight containers are well suited to them, although the industry will continue to fight over the definition of "grid" since Oracle and IBM have widely divergent approaches.

6.  The battle for social networking prominence will be played out in 2008 as MySpace, Facebook, Plaxo, and LinkedIn position themselves for enterprise use. As the "IM generation" enters the workforce, they are going to expect support for familiar social networking technologies in the corporate environment, encouraging corporations to figure out how to incorporate them into business culture, but one or perhaps two winners will emerge from the battle in 2008. Meanwhile expect employees to hedge their bets by taking out pages on multiple sites, causing confusion in the short term over which site to favor...

Semantics • Event-Driven Programming • AJAX Consolidation • Flex vs AJAX
Bill Roth
Vice President
BEA

Bill Roth is a member of the editorial boards of both Java Developer's Journal and WLDJ. He is vice president of the BEA Workshop Business Unit. Prior to this he was chief technical evangelist for Epiphany. With over 19 years in this industry, he has played numerous product marketing, product management, and engineering roles at companies like Sun and Morgan Stanley.

In my view i-Technology is heading in a few discernable directions:
1.  Semantics: The next quantum leap in computing will be in the area of annotating information with additional meaning, i.e., semantics. Tim Berners-Lee saw this in 1999 when he wrote about The Semantic Web. The idea is that if you augment data with additional information that allows a computer to determine what it actually means, you will be able to do more with that data, and be able to take more human processing out of the loop. Semantics is not a new topic. Researchers in the 1960s made their first stab at it. But with the advent of RDF and OWL, we may be able to achieve the first tangible (and commercializable) improvements in computing.

2.  Less coding: The essential productivity of the programmer has not improved in 25 years. The fact remains that it is still difficult to write code. As a result, more and more systems will become available that make it possible to build applications with less code. Metadata-driven development will also start taking hold in the next 12 months.

3.  More programming paradigms, like event-driven: Because of #2, more fit for purpose programming paradigms will emerge. The most likely candidate in this area is "event-driven" programming, a mélange of declarative and rule-based concepts for building our application in what Gartner calls Event-Driven Architecture.

4.  Consolidation of AJAX models: There seem to be hundreds of AJAX programming models, both open source and commercial. The coming 12 months will see a consolidation to four or five models. No one model will become dominant yet. Furthermore, since the browsers were never really designed to solve the RIA problem, there's a good chance that alternative models such as Flex and Silverlight will make a serious challenge to the AJAX models.

5.  Flex as an alternative to AJAX: While AJAX has hundreds of models, Flex has but one. Built on the widely deployed Flash technology, Flex has easy-to-use tools and a powerful scripting model that allow amazing things to be built. The simplicity and deployment of this model will prove to be very compelling in the enterprise in the next 12 months, and will make Flex a serious rival to all of the AJAX options. (Disclosure: I shamelessly admit I am a huge fan of Flex.)