Last week, I discussed some major product releases we can expect from Microsoft in 2005, and the product list garnered several questions. As the final year before the company ships Longhorn, the next Windows release, 2005 should be a transition period without many major product releases. However, the release of Windows XP Service Pack 2 (SP2) caused many product updates to be pushed back, and now we're going to see a busy 2005 release schedule. This rush of product releases is entertaining for me, of course, but frustrating for those who have to plan rollouts.
In 2004, Microsoft pledged to better communicate its enterprise product release schedule. The company then proceeded to make at least two major changes to that schedule later in the year. Now, Microsoft's scatterbrained approach to enterprise product shipments is raising some serious questions with customers.
Software Assurance Questions
A major concern involves customers who opted for Software Assurance (SA) volume-licensing agreements. SA customers receive free upgrades to new software versions during the lifetime of their SA contract. But during the lifetime of the first SA contracts, at least two major products have been delayed significantly: Microsoft SQL Server 2005 and Windows Longhorn. Now it appears that because of these delays, some SA customers won't receive the promised software upgrades. Unless Microsoft does something to rectify this situation, the results could be interesting. Some analysts have suggested that customers will abandon Microsoft's SA program in droves. I'm not sure that will be the case: Enterprise customers typically make changes slowly--whether it's to adopt technology or drop it. But the aftereffects of the bad feelings this problem will cause are hard to measure.
Given the heavy security focus of Windows IT Pro UPDATE commentaries in 2004, I don't think I need to belabor the point here. Microsoft's products are under attack every minute of every day from malicious hackers and dubious online entrepreneurs hoping to find unprotected PCs and take advantage of the hundreds of security vulnerabilities that continue to dog Windows PCs. I'd rather write about anything other than security, but it's a fact of life and unavoidable.
Some of our current security problems are self-inflicted, but Microsoft has rightly assumed the burden and is moving, if slowly, to a more secure product line. XP SP2 was the first major step toward that goal, and Windows Server 2003 SP1, due sometime this quarter, will bring that pervasive technological protectionism to the server as well. But the underlying security concern is an age-old one. Just a few years ago, we debated whether Microsoft's tactic of selling us on the security updates that are available only in the latest product versions was a good practice. Today, this fear is a reality: If you want the most secure Microsoft-oriented infrastructure, you'll want XP SP2 on the client and Windows 2003 SP1 on the server. For many companies, large and small, that's an unobtainable goal. Perhaps Microsoft could provide financial incentives for making the upgrade.
And Microsoft's security strategy has technical holes, some of which are just now being filled. The company bought GIANT Company Software recently for its excellent antispyware application, which Microsoft is providing to the public free as Microsoft Windows AntiSpyware (Beta). However, it's a client product and not easily deployable in large companies. Sometime later this year, I expect Microsoft to ship an enterprise version of the antispyware product. I hope to have more information about an enterprise antispyware product soon. Also, today Microsoft will ship its first malicious software (malware)-removal tool along with the January 2005 monthly security fixes. This tool will remove viruses, worms, and other unwanted electronic attacks but likely won't provide any real-time scanning capabilities. I'll provide more information about the tool as information becomes available.
I've been testing x64-based software, including prerelease versions of XP Professional x64 Edition and Windows 2003 x64 Edition, since the third quarter of last year. I've been generally impressed with x64's quality, performance, and compatibility. Out of the box (so to speak, there won't actually be a box), XP Pro x64 includes better driver support than did the original shipping version of XP 32-bit. And as we've gotten closer to the XP Pro x64 release, more and more applications are starting to work properly, providing a near-seamless XP experience.
That said, I think I've identified a major problem with XP Pro x64, and it's going to hamper migration. There are still far too many applications that won't run on XP Pro x64 and many that won't even install on the platform. Two main reasons for the latter problem exist. First, many applications (including a lot of Microsoft applications) perform a version check and incorrectly assume that XP Pro x64 is a future version and refuse to install. Second, many 32-bit applications still inexplicably use 16-bit installers. Because x64 technologies no longer include a 16-bit subsystem, 16-bit installers can't run. Third, entire classes of applications don't even exist in x64 format. For example, there are no tier-one antivirus applications that will work on XP Pro x64 yet.
Then there's the driver situation. Although XP Pro x64 ships with a great collection of stock drivers, you're out of luck if any of your drivers aren't included in the distribution. I have a wide range of hardware in my home office, including two recently purchased scanners, which refuse to work with XP Pro x64. The driver situation, coupled with the application problem, will bite many hopeful users. Although x64 is clearly the platform of the future, I'm beginning to think that people might be better off skipping XP Pro x64 for the short term.
If you find the x64 situation unsettling, at least you didn't standardize on Itanium hardware. I got a nasty-gram from someone at HP about my comments last week about the Itanium. But I stand by my original statement: In my opinion, the Itanium is a technological dead-end, albeit one with a short window of viability in certain high-end scenarios. But Microsoft and the industry are rallying around x64, not Itanium, and are supporting x64 with a wide range of products and servers. Although the x64 platform still has a way to go to achieve the scalability and performance of Itanium-based servers, it will happen and probably will happen quickly. Therefore, Intel's plan for a single 64-bit hardware platform that can hot-swap between both x64 Xeon chips and Itanium chips, seems rather quaint, especially when you consider that the platform won't ship until 2007. 2007? We'll all be running on x64 by then.