The best desktop strategy: Forget the desktop
Before licensing and rolling out an updated operating system, it may be time to revise your IT estate
Windows 10 is already running on approximately 100 million devices.
As Microsoft is offering Windows 10 as a free upgrade for users from the decent but old Windows 7, and the badly implemented and generally disliked Windows 8, consumers have been taking the opportunity to move to the new, continuous-delivery version of Windows – the last that Microsoft will provide a number to. From now on, new functionality will be added at regular intervals, downloaded at the same time as the regular patches.
For businesses, this raises a few issues. The end of life of the Windows XP operating system (OS) has been much discussed, with those still using it finding support increasingly expensive – and open to security risks and a lack of a native, up-to-date browser. Organisations on Windows 7 and 8 are wondering whether the licence cost, (Windows 10 is not free to businesses, but is covered under volume licences supporting upgrades), and the business cost, (the upgrade work, re-educating users and retraining help desk staff), of the change are worthwhile.
It is probably a good time to completely review where your organisation is with its user device strategy. As bring-your-own-device (BYOD) programmes have come into broader use, users are no longer necessarily using Windows-based devices as their main mobile appliance.
Apple iPads and Google Android tablets have gained the lion’s share of the market in the mobility arena – and these do not run Windows applications natively. Microsoft may have released native versions of its Office suite for tablets, but few other Windows application suppliers have gone to the trouble.
Some commentators have long predicted the fall of Microsoft, with Linux as a desktop OS taking over, and free OS software replacing commercial off-the-shelf software. This has not happened to any great extent – the support skills and need for user retraining on using a new OS and supporting new applications is just not very attractive.
Most still use old Windows applications – one of the big reasons many have stuck with Windows XP. After XP, Microsoft changed the way applications were run – and 30% of applications written for XP did not run directly on later OSs. Vista was released in 2007 and any applications you have that would not run on Vista are now likely to be over 10 years old – and due for a thorough review as to their suitability for purpose.
What is the desktop for?
If you do decide to stick with Windows and migrate to Windows 10, then PC suppliers such as Dell, HP and Lenovo will have the capabilities to help with migration – and to put in place the capability to run old applications on the new platform, either by creating a new installation method that will create a working application (as developed by ChangeBASE, now part of Dell; or AppDNA, now part of Citrix); or to run a virtualised environment on top of Windows 10 for a specific application.
However, many may be starting to wonder just what the purpose of a desktop is, anyway. It used to be a self-contained computer used for running local applications and possibly saving the data created on a shared platform. Is that really the case any longer?
With the increasing use of software as a service, a lot of workloads are run on servers in a facility under someone else’s control – all the user sees is the graphical interface. Therefore, desktop performance is no longer the main issue; indeed, Intel has realised this and we will see far less in the way of business messaging around how fast desktop machines are, other than for workstations used in areas such as engineering or media work.
This will still leave a lot of applications in business that need a Windows platform. For those who are mobile, installing these onto a discrete desktop machine doesn’t make much sense – they will only be able to use the applications when they are sat at that machine. Issuing them with a laptop and a parallel set of applications can lead to licensing and data issues – and they will probably want to use their Apple or Android device anyway. So what to do?
The obvious route is to look at a virtual desktop infrastructure (VDI). By concentrating all desktops in a central datacentre, organisations not only provide a consistent experience to users across multiple different devices and device types, but also regain control over the data users create.
The main suppliers in the VDI market are Citrix and VMware, although a large ecosystem of other suppliers has grown up to add greater capabilities. For example, RES Software and Centrix Software provide discovery tools that enable organisations to see who is using what before they migrate, to understand what the best desktop systems would be for groups of people, and roll the desktops out in a manageable way.
FSLogix provides software that enables images to be created that have strong control over what users can run and what they can install themselves through virtualising the Windows registry, minimising the need for multiple different server images.
VDI provides the capability for users to access their Windows-based desktop from their non-Windows tablets and smartphones – but this can lead to performance issues. Numecent provides an approach it calls “cloudpaging” that can harness the power of the end device to run Windows-based applications at native speed. Applications too slow to run over the wide area network in a VDI environment can be streamed down to the device and run as a native application, providing a much better user experience.
Read more about desktop IT strategies
A VDI pilot project should start with a VDI project plan. Know what pitfalls to avoid and test product options to achieve a successful VDI implementation
Many organisations skipped the PC-unfriendly Windows 8 release. But Windows 10 is far more enterprise-focused
Days are numbered for desktop as a workhorse
In most cases, the desktop and other user devices are rapidly moving from being native devices running enterprise applications (even in client/server mode) to a pure access device, enabling users to mix and match between Windows, non-Windows and cloud-based apps and functionality, through browsers or thin clients. The capability of HTML5 to create a strong user experience should not be overlooked – creating applications that can run directly to any device is becoming far easier.
This can offset the cost of VDI implementation by running existing PCs for longer, sweating the assets. In some circumstances it will work but, unless a third-party browser is installed, much web-based capability will not be available with older Internet Explorer versions, and will leave security vulnerabilities that can be exploited by hackers, for example to infiltrate the VDI servers.
Intel has decide that, in its revised approach to the desktop – where it knows it cannot push speeds and feeds any more – it will build security directly into the processor. Multi-factor authentication is used – yet the signature created by a fingerprint scanner, iris scanner or other biometric device is stored in the processor and cannot be directly read. Instead, when using biometrics, a signature is created and compared to that held on the chip. If it matches, the chip issues a security token used to provide access. When the device shuts down or goes into sleep mode, the token is revoked, to be issued as new next time.
Such an approach means users will no longer need username/password pairs – so improving security while reducing the need for organisations to have large numbers of agents on helpdesks helping users reset their passwords.
Through this and other changes to the capabilities of desktops, Intel hopes it will still be able to drive the need for organisations to refresh their desktops – but whether or not those organisations will regard such capabilities as retaining sufficient value to the business for broad-scale refresh programmes to be carried out remains to be seen.
So, yes – it is time to review what is happening at the desktop. It is also time to forget the desktop and concentrate on what the business wants from its end-to-end IT platform. In most cases, this will lead to an understanding that a desktop PC is no longer what is required as a workhorse – although it may still be useful as a simple access device.
Clive Longbottom is founder of analyst Quocirca.
