Ten Emerging 2005 Software Trends

Only a hermit would not have noticed the massive, and I emphasize massive, developments in the Software development world in the first half of 2005. To get a handle on the most recent developments, I consult back to my original prediction as a frame for the picture.

Firefox Browser market share will continue to grow in 2005. The size of its growth continues to be a source of debate, however without a doubt it has made progress and in ways that would have not been predicted. The inclusion of "integration at the glass" features and the combination with social networking applications have drastically changed the rules of the game.
Eclipse marketshare will continue to grow. Eclipse juggernaut has shown that resistance is indeed futile with the assimilation of BEA, Borland and Macromedia. The error in my original prediction was that Oracle would become more active participant rather than Borland. Borland's change of heart was a surprise, however there's no better convincer than the market.
AMD will continue to increase its performance dominance over Intel. I didn't expect dual-core processors to be introduced and then delivered so quickly. However, recent tests at Tom's Hardware have shown AMD's dual cores increasing their performance (and reliability) lead over Intel based alternatives. AMD is exceedingly being adopted as the preferred processor for massively parallel machines.
Javascript will regain dominance in the space of Rich Internet Applications (RIA). XHR has driven massive inovation in web application development. The term AJAX is becoming exceedingly all too common. The ramifications of technology affects how code is written in both the client-side and server-side. Code at the server side is exceedingly looking like code for delivering data independent of presentation. The difference however from a pure database solution is that the data is delivered in an implementation agnostic manner. Another, interesting observation is the importance of ReSTful principles in this new kind of architecture.
Java developers will continue to abandon EJB as the standard way of building Java based server applications. Lightweight frameworks like Spring will continue to take marketshare. Tim O'Reilly in a recent blog entry reveals the re-emergence of interest in Java based technologies, particularly Open Source based technologies. The Java API space continues to be fragmented, however there is an emerging trend of adoptin Spring as a basis of a meta-framework. There have been earlier forays into building a "framework to bind all frameworks". However none have been quite successful until now. That's because Spring adoption is happening in a more decentralized manner.
Semantic XHTML will continue to gain mind share as the best way to encode semantic information. A new blog (www.microformats.org) has been created to track this space. There's a growing interest in this approach, however the verdict isn't in yet. The semanticweb folks have taken big strides in the first half of 2005 fending of attempts to simplify the RDF syntax.
SOA will continue to mesmerize the masses... WS-* will move close to irrelevance despite the huge marketing campaigns of many vendors. Clearly, innovation is developed elsewhere. The latest development in this space has been to refactor WS-* to more simple and ReSTful forms. However, there's no easy way of avoiding complexity by virtue of missteps taken during evolution.
Large costly multi-cpu servers will be abandoned in favor of lots of cheap low cost single CPU servers. From the economical sense the viability of this approach is becoming increasingly clear. Unfortunately, what is lacking is architectural guidelines and frameworks to develop and manage such a system. Azul systems has released a massively parallel java processor that can be leverage with minimal code rewriting. Pillar Data Systems has a scalable storage solution based on low cost components. Terracotta has a transparent clustering solution based on AOP technology.
IT Deflation will be the continuing trend. The kids are apparently shunning computer science education, and I've been making less money than I made back in 2000.
Scripting languages will become dominant in addressing the needs of situated software. The next version of Java (code "Mustang") will include a Javascript interpreter based on Rhino. A case of creaping featuritis or a neccessary enhancement?