In January I published a set of technology predictions for 2005. 6 months on, I am reviewing those predictions.
I posted the first and second parts of the mid term review a few days ago.
This is the third part. Ostensibly there is only one more prediction to review, but I decided to add another two...
Apple To Prosper
My original article stated:
"Apple's PC market share is helped to a certain degree by the security woes that Microsoft is suffering — but there is clear consumer enthusiasm for the iMac too. We expect to see this strengthen in the coming year. However, we do not expect Apple to make any significant in-roads to the corporate PC market. The legacy PC technology provides too much of an obstacle for it at the moment."
The consumer enthusiasm for the Mac has grown more strongly than I suggested or expected. Market growth for Apple PCs now seems to be around 70% or so. This is primarily growth in the home market (as expected), boosted by the fact that a significant number of iPod users are buying Macs.
The Apple switch to Intel deserves comment. Comment has suggested that this is due to Apple disappointment with the G5 (particularly IBM's inability to deliver a G5 for Apple's laptops). It may be a factor, but I doubt it.
What I believe is this:
Steve Jobs looked at the PC market (Dell, HP, Sony, Lenovo, Acer, etc) and thought:
"None of these companies innovates in any significant way. Apple does (and not just in software). We can beat them. If Apple runs on Intel, you can load Windows (in a partition) and you can load Linux (in a partition), but it will still be an Apple booting OS X. Nobody else will be able to do that. We can have a much bigger market. More developers too."
And also there are two chip manufacturers that can compete for Apple's business.
So I offer a new (and very early) prediction. Apple is going to change the dynamics of the PC market — its dramatic growth rate will persist once it moves to Intel. Watch it happen.
Event Stream Processing (ESP)
I'm adding two new trend predictions here as a consequence of what I've observed in the past six months.
The first concerns streaming—or ESP as it is being called. Earlier this year a new start-up, StreamBase, came to market with a new type of software product (now in version 2.0), which focuses on the real-time processing of data streams. Very soon afterwards, Progress Software acquired Apama—a streaming development environment—to complement its ObjectStore database. This created a second product that was also targeted directly at the processing of streams.
In my view these two developments created a new product category: ESP Development Products.
I expect both of these products to enjoy success and I also expect other vendors to enter the market. Indeed, I expect the real-time processing of data streams to become an expanding area of software activity for many years. In my view it is a new category of IT activity—a little like data warehouse was when it first came on the scene. In other words it is a whole technology area, around which a variety of capabilities will be created.
At this moment it is an infant market and there are only a few industry sectors that have obvious areas of application (financial services, telecoms, defence, security and one or two others). In time, however, virtually all organizations will have areas of application (RFID will ensure that).
Right now, TimesTen—which was recently acquired by Oracle—has started to position itself as a player in this market. I'm not so sure it qualifies, but I haven't looked at it in a while so maybe it has changed. It used to be an in-memory accelerator. ESP products are more than that. They have to be fast development environments and they have to be built to manage data streams.
SOA & Web Service: "Sex with Aliens" No Longer
Here, I'd like to update an observation from January 2003 when I wrote "Web Services is currently like sex with aliens: some people say it happens, but nobody can furnish any proof."
Well 2 years makes a big difference. The reason for Web Services skepticism at the time was that the technology was in its early hype phase and the problems of security had not been addressed. Nor had the problems of management. So Web Services applications tended to be confined to Intranet applications. This has changed. Now I'm seeing a fair amount of Web Services applications deployed and am running into more and more companies that are embracing SOA. The embrace of SOA is almost total among the vendors — not just development tool vendors, but vendors of almost all kinds of software.
This is important. A kind of software development consensus is emerging.
The point to note, or trend prediction if you like, is that SOA is now dominant and that from here on (at least for the next few years) software development will focus on the issues of SOA (although we'll have to allow for the fact that there is still a good deal of misunderstanding in the community at large as to what SOA actually is).
Focus points will include the problems that have not been solved (the problems of security, operational management and life cycle management of SOA are still to be solved and will become increasingly visible issues) and debates on what is and is not the best approach to development.
Expect to hear a good deal talked about business friendly development environments. (The term BPM will become less meaningful as time goes on).
Expect also to see too little attention placed on the problems of legacy maintenance, which are not yet solved by any products in the SOA space.
