Before I go any further, let me set one thing straight… for the majority of Microsoft’s enterprise customers, there are very few reasons why Windows XP should be deployed on new PCs in preference to Windows Vista.

Vista has now been available to enterprises for over a year and a half, has long since passed the first service pack release, and many of the initial difficulties are now resolved. It’s not perfect, but very few software products are. For that matter, neither is Windows XP (nor for that matter are Mac OS X or the various Linux distributions).

Sadly, for Microsoft, the general perception of Windows Vista is not a good one. Ask anybody who ran Vista from day 1 and they will have stories of the problems that they have had because ISVs and IHVs were too slow to update their applications and device drivers (hey, they only had 5 years notice…). But ask the same question from anybody who waited a few months and it’s a different story – Vista runs perfectly well on most modern PCs (and I don’t mean an exotic machine with a fantastic custom hardware specification – I mean pretty much any business PC purchased in the last few years, as long as it has enough memory). Unfortunately there is a saying that perception is reality.

So today is the last day that you can buy Windows XP. Except that it’s not really. In an open letter to Windows Customers entitled “An Update on the Windows Roadmap”, Bill Veghte (Senior Vice President for Online Services and Windows Business Group at Microsoft) explains that:

“It’s true that we will stop selling Windows XP as a retail packaged product and stop licensing it directly to major PC manufacturers. But customers who still need Windows XP will be able to get it.”

So you can get XP. But why would you? The simple fact is, that if you are looking to deploy a new Windows desktop in the next few months, then basing your plans on XP is building a problem for later.

In terms of the product cycle and future roadmap for Windows XP (Professional, 32-bit):

• Support for Windows XP RTM and SP1/1A has already ended.
• Support for Windows XP SP2 will end on 13 July 2010.
• Although no official announcement has been made by Microsoft, it seems unlikely that there will be a fourth service pack for Windows XP. On that basic, mainstream support for Windows XP SP3 will end on 14 April 2009 and extended support (i.e. security patches only) will be available until 8 April 2014.

Some of my recent customers are just starting rollouts based on Windows XP SP2. By the time they have completed their rollouts, XP will be on extended support. And if they don’t move to SP3 soon, then they will be unsupported. In most cases the reason they are not considering Vista is application compatibility – in which case they should really be looking at application upgrades, or possibly using desktop/application/presentation virtualisation technologies – but why does a move to Vista have to be wholesale? A co-existence strategy involving managed diversity on the desktop is the way forward for many organisations.

So, I’ll finish up by repeating what I said at the head of this post:

“For the majority of Microsoft’s enterprise customers, there are very few reasons why Windows XP should be deployed on new PCs in preference to Windows Vista.”

Additional reading

• Windows XP: Facts about the future.
• Windows Vista: Look how far we’ve come.
• An Update on the Windows Roadmap

Many people expect Windows Vista SP1 to be a turning point for deployment of Microsoft’s latest desktop operating system. Critics have derided Vista, citing it as a disaster for Microsoft, suggesting that it has suffered from adoption rates. Others say that the desktop operating system model is a thing of the past – to be replaced by a “webtop” of cloud-based services. I don’t think that day is here yet – and anyway, we’ve been here before – weren’t thin clients supposed to have taken over by now?

But has Vista really done that badly?

Journalist Paul Thurrott used an interesting comparison in a recent podcast (Windows Weekly episode 48). In a market where 260 million PCs were sold worldwide, Microsoft sold 100 million copies of Vista. Some of those would have been downgraded to Windows XP but what about the 160 million PCs sold without Vista? What were these running? A few Linux machines, some Mac OS X, but mostly Windows XP – and this is what some people are highlighting as a failure to sell Vista.

Thurrott was unable to locate global PC shipment figures for the same period after the Windows XP launch, but he used another method to compare the adoption rates of the two operating systems:

• At the time of the Windows XP launch, the global installed base of PCs was around 250 million and Microsoft sold 23 million copies of XP in the first year. That equates to an adoption rate of around 9.2%.
• Looking at the figures for Vista, on an installed base of around a billion PCs, Microsoft shipped 100 million copies – an adoption rate of around 10%.

So, Vista has had a greater adoption rate than XP in its first year and, however you look at it, Microsoft sold 100 million copies. To my mind, that’s not stunning – but not bad either – and it seems that Windows Vista adoption is on the increase. CDW’s Windows Vista Tracking Poll (January 2008) suggests that:

• The number of organizations evaluating and testing Windows Vista has increased to 48% in January 2008 (up from 29% in February 2007 and 21% in October 2006).
• 30% of organizations are currently implementing or have implemented Windows Vista.
• Windows Vista is delivering on expected benefits, with nearly 50% of evaluators/implementers reporting performance above expectation on key features.
• And, although not part of Windows Vista, but of equal significance whilst examining adoption of Microsoft’s core technologies, 24% of organizations have implemented the Office 2007 System, up from 6% in February 2007.

All of this suggests that Vista (and Office 2007) are already pretty successful. So why the perception that Vista is not ready for the enterprise?

The first barrier is the “wait for the first service pack” mentality. Regardless of its validity, this view certainly exists and the release of SP1 may allow some organisations to start their preparations.

Other perceived barriers are the hardware and software requirements for the operating system but the reality is that any system purchased in the last few years should be capable of running Vista. And, when it comes to device drivers and application support, Microsoft is caught up in a vicious circle where vendors are reluctant to invest in updating their product to work with Windows Vista and customers will not deploy the new operating system unless their hardware and application software requirements can be met. This is the reason that, according to Paul Thurrott, Microsoft worked to ensure that SP1 will resolve issues for 150 enterprise applications that were blocking large-scale customer deployments.

The third issue that I see is that of cost. In the late-1990s, we saw organisations perform technical refreshes every three years or so, fuelled by a combination of technology advances and preparations for avoidance of the “millennium bug”. In recent years, however, the need to roll out the latest and greatest has been tempered somewhat. Rolling out separate hardware and operating system upgrades is often seen as double the trouble, and, unless there is a business benefit that exceeds the disruption and cost of a new desktop environment, organisations are slow to make changes.

Instead, many organisations are considering a system of managed diversity – running Windows Vista on new (and recently purchased) systems but sticking with XP on older machines that do not yet warrant replacement, or where applications do not yet support Vista. This was what Gartner recommended back at the time of the Vista launch and it’s for exactly this reason that I have been critical of Microsoft for taking so long to develop a third service pack for XP – by the time SP3 arrives it will have been almost four years since the last one.

Finally, there is the issue of new features. Windows, like Mac OS X, and any other mature operating system has reached a point where some think it has too many features and others say it needs more. Microsoft has a particularly difficult battle, whereby if it bundles software with the operating system it falls foul of competition laws. It seems to me that many of the Windows Vista improvements are incremental – and that makes a wholesale migration difficult to justify. Perhaps the strongest argument to date has been productivity improvements but these may be offset by people needing to learn new methods of working. With the release of Windows Server 2008 this will start to change – new technologies like network access protection and some of the networking enhancements require a new server infrastructure and that’s when we will start to see a stronger case for adoption of new technology.

Microsoft’s problem is persuading customers to make the move from its own legacy and even when Windows XP is withdrawn from sale in June 2008 (although system builders will still be able to provide XP pre-installed until January 2009), extended support will continue until 2014. Interestingly, Gartner, the same organisation that advised customers to wait before moving to Vista is reported in IT Week as warning firms to start the introduction of Vista no later than 2009 because software vendors are likely to start phasing out Windows XP support after this.

Service pack 1 for Windows Vista is now available for customer download (with some restrictions). It won’t be released on Windows Update for a few months, due to issues with certain hardware devices for which new device drivers will need to be released first, but for those 48% of organisations that are evaluating Vista, SP1 will play a major part. Further details of Vista SP1 and its release schedule may be found in Paul Thurrott’s Vista SP1 FAQ.

Windows Vista may not be perfect (no desktop operating system that I am aware of is) but it does offer improvements over its predecessor and is reaching the mainstream business market. SP1 will accelerate the adoption rate but the main change is that, for many organisations, the move to Vista may be a gradual one and strategies for managing co-existence with legacy operating systems will be crucial.