Todd's World: Windows XP, Vista, 7…what should my desktop strategy be?

Windows XP, Vista, 7…what should my desktop strategy be?

Published 30 January 09 04:47 PM

So, the past several years have seen some interesting developments in the Windows client lifecycle. Many customers are taking a hard look at their current environment, the overall technology landscape, and the Windows roadmap. Now is a good time to try to determine what the best course of action is with respect to the Windows platform. To that end, I thought that it would be prudent to put some thoughts in writing – some of them the official corporate stance, some of them my own personal spin. Here we go…

Windows 2000 has largely gone by the wayside on the client – thankfully so, as support for 2000 ends on July 13, 2010. Anyone who is convinced that we’ll extend that date is playing with fire. Yes, we did so with NT4, but that shouldn’t be considered a precedent setting event. Retire those machines this year so you don’t have a firedrill in 2010 – especially if your company is subject to SOX, PCI, etc.

Windows XP is still going strong – the most popular version of Windows from a deployment perspective, with extensions to the OEM licensing deadlines and introduction of lower cost netbooks having an impact. That being said, there’s definitely a level of comfort out there in the enterprise for Windows XP. It’s been around for over 7 years now. ISV’s and OEM’s have been developing applications and drivers on this platform for a long time, and business users & consumers alike are familiar with the interface. At the same time, let’s think back to what was happening in or around 2001: Apple introduced the iPod and OSX, Google had just 8 employees two years earlier (they have almost 17,000 now, if you were wondering), 3G cell service was first launched (in Japan), a 1 Ghz processor was very fast, and you almost undoubtedly used the modem in your laptop with great frequency (and maddening slowness). Times have changed, but you’re probably still running the same version of Windows that became available back then. Imagine if you were still using a first generation iPod or had a 1 Ghz Celeron under your desk! Of course, moving music to a new device is different than upgrading an operating system, but you get the idea.

Service Pack 3 was released last spring, and as a consequence, SP2 support will be retired on July 13, 2010. Put that date in your calendar and consider it D-Day for legacy Windows! Windows XP leaves Mainstream Support and enters Extended Support in just a few months - on April 14, 2009. For the full breakdown on what this means, I encourage you to speak with your Technical Account Manager, if applicable, and\or visit the front page of the Microsoft Support Lifecycle site: http://support.microsoft.com/lifecycle/

Windows XP with SP3 will be supported until April 8, 2014. So that gives XP environments some breathing room, but not necessarily as much as you might think (more on that in a minute).

Windows Vista is a loaded topic. No one that I’ve met feels ambivalently about it, for one reason or another. Admittedly, the launch did not go as well as we would have hoped. We made huge changes ‘under the hood’ that made it difficult for driver and application developers to ramp up – as a result, the Windows ‘ecosystem’ was not prepared. Vista itself had some problems with performance and stability that were fairly well publicized (and the Mac vs PC ads definitely didn’t help)…so 2007 was a rough year for adoption. The beginning of 2008 saw the release of Service Pack 1, and a years time had given our partners a chance to get their hardware, drivers, and applications updated for the latest operating system release. So, the folks who have seen and used Vista over the past twelve months probably have had a good experience with it. We’re finding that the major problem out there today is public perception. Go to http://mojaveexperiment.com/ if you don’t believe me. All that being said, enterprise adoption is about what we saw with Windows XP 2 years after it’s release. The one recurring theme in discussions with corporate customers is that application compatibility is a problem. Applications may not run in Vista, or maybe they can, but it’s not supported by the vendor. Remediation will be costly and time consuming. We get it. Many of the acquisitions and investments we’ve made in the past few years are targeting that problem specifically (Application Virtualization – SoftGrid, Enterprise Desktop Virtualization – Kidaro, etc.) Operating systems are traditionally tied to hardware, user data, and applications. We want to decouple them so that it is feasible and relatively easy to perform an in-place OS migration. Our Desktop Optimization Pack technologies are a must have for those considering an upgrade any time soon.

On the positive side, there should be no one out there that can deny the security enhancements Vista brings to the table. It’s far and away the most secure OS we’ve released to date (the first time that we’ve gone from drawing board to release with security as priority #1). Mobility functions and power savings features are also prime examples of the benefits that you can reap from Vista. The 64-bit release is very solid, and people should be looking in that direction, as memory grows cheaper and cheaper. Any modern hardware (2-3 years old) with 2 GB of RAM or more should run Vista quite happily.

The hot topic. The Windows 7 beta release several weeks ago was met with overwhelming interest (we couldn’t handle the initial flood of download requests). The reviews are almost all positive – even Walt Mossberg of the Wall Street Journal had good things to say. And he’s not, ahem, usually all that much of a Microsoft fan. Millions of people are out there running it on thousands of different hardware configurations, and I’ve been running it myself on laptops, desktops, and server class machines with minimal hiccups. The caveat to this early success is that we’re hearing from a lot of folks ‘Why should I upgrade to Vista when Windows 7 is right around the corner?’ Well, the answer to that is what I’ve been building up to. If we look at it from the perspective of an enterprise with fairly unaggressive adoption cycles, then you’ll see that you may be putting yourself in an untenable situation a few years down the road. For the sake of argument, make these assumptions:

Company A doesn’t deploy new operating systems or major applications until Service Pack 1 (or a similar bug-fix milestone) has been provided by the vendor

Company A probably won’t even begin testing their application footprint against the new OS until said SP1 is available

Windows 7 ships in the fourth quarter of 2009

Service Pack 1 for Windows 7 would likely not be final until the first half of 2011, if not later (going by our historical timelines for SP1 releases)

So, Company A would begin testing migration from Windows XP to Windows 7 SP1 in 2011 sometime. How long would it take to perform adequate testing of your application suite to certify\remediate it for Windows 7? For most, this is at least a 6 to 12 month process…so, now we’re in mid-2012. At that point, you’re ready to start building an image (hopefully using the MDT to make your lives easier). Maybe the image is ready to go in early 2013. Then you have a little over a year to get it out company-wide until Windows XP hits end-of-life. Is that enough time? Perhaps…but is it worth backing yourself into a corner?

A few points to make in conclusion. Thanks for sticking with me this long. :)

While Windows 7 is a major release from our perspective, it does not involve any huge architectural changes, so I personally feel that the ‘waiting for SP1’ methodology is unnecessary. Given how aggressively that we’ve been working with the partner ecosystem, I would expect that they will be much more ready to support Windows 7 at the time of availability than any prior release of Windows. Based on the timelines that I described above, I would also argue that it’s a greater risk to sit on XP and wait for Win7 SP1 than it is to go to Vista and\or Win7 RTM.

One of the primary tenets of Windows 7 development is maintaining compatibility with Windows Vista applications and drivers.

Starting to test against Vista now will bear fruit in the long run – you won’t have to perform two runs of testing on Vista and Win7

Deploying Vista now to at least some subset of systems, will provide you with experience that will be valuable for Windows 7, and take advantage of the much improved security, mobility, and power-saving features in Windows Vista today

The UI change going from XP to Win 7 will be greater than going from Vista to Win7. I know that user training is a very important topic to a lot of folks.

Start to investigate desktop virtualization now. If you have a significant portion of desktops (as opposed to mobile PC’s), then a VDI scenario warrants serious consideration. We have partners such as Quest and Citrix that have solutions in this arena that make it a reality *today*. I’ll be writing up a Windows VDI post soon to cover technology and licensing requirements.

Analysts such as Forrester and Gartner are already on the record as saying that customers should not plan to skip Windows Vista

The Desktop Optimization Pack technologies are going to make OS migrations much, much easier than what was dealt with in the past

The Microsoft Deployment Toolkit (Formerly the BDD) provides the tools necessary to build and customize images very easily, with simultaneous administrator ease-of-use and handling of complex scenarios

All in all, please remember that Microsoft is here to assist with your OS migration and testing. Whether you would like someone to help get your organization going with the MDT, application compatibility testing, Windows Vista or 7 readiness – we have resources that can help (at no cost). Give me a shout to get the process started.

by mfiorina