Who can explain desktop virtualisation?

When it comes to desktop virtualisation, there's too much hype and not enough explanation. That's the message coming across loud and clear from sources as diverse as a user survey conducted by a vendor, and from an analyst specialising in virtualisation issues. It all sounds wearyingly familiar.

On the one hand, a press release from Fujitsu suggests that "Industry over-hype is preventing CIOs from achieving the benefits of desktop virtualisation and halting the move towards a fully virtualised desktop infrastructure. Research from Fujitsu has shown that CIOs across the private sector lack the clarity and conviction around desktop virtualisation, with the majority (49%) of them believing that desktop virtualisation technologies are promising more than they can deliver," it says here. It goes on to say that "desktop virtualisation is over-promised", a view held especially by mid-sized companies.

On the other hand, a piece by analyst Tony Lock of Freeform Dynamics in Another Place suggests that vendors simply aren't doing enough to explain the benefits of desktop virtualisation -- or more precisely, what the various options are along with their respective advantages and disadvantages.

Instead, we just get the marketing departments' usual stuff about how "ours is the latest and greatest and you've got to buy it NOW".

Here's a vignette. I was in discussion yesterday with a vendor of a perfectly respectable-sounding service (I've not been able to try it so have only his word for it) that promises to take the hassle of managing PCs off your hands by providing virtualised desktops. It sounds flexible, insofar as you can access it from a range of devices and even host it yourself if you've the infrastructure, easy to manage, and runs using industry-standard software, such as Sun and Citrix.

I won't embarrass him by naming the company as this is such a common phenomenon but, when I asked about issues such as SLAs, objections or challenges that the service might have or which customers had raised, or who the competition was, said vendor was unreassuringly vague. To be fair, his justification was that the number of failures was so low meant that an SLA's break-fix model wasn't appropriate, and did eventually come up with a list of competitors -- but these are simple questions a customer would ask.

Why not be open and realistic? Customers would surely welcome that.

It all suggests that desktop virtualisation is a technology that has some way to go before it climbs out of Gartner's infamous trough of disillusionment and heads for the plateau of productivity.

However, I'm convinced that desktops as a service is the right way to go -- especially if you can still run fat clients. When the thin client mania first hit back in the 1980s, it was the terminal vendors who shouted loudest as they watched their revenues dwindle before the onslaught of the PC desktop. Their arguments revolved around issues such as centralised control and ease of management. Those arguments for thin clients are the same today but the easy deployments have already been made, for employees such as call centre workers and other single-taskers.

Experience suggests that you won't convince people that thin client-only model works in all circumstances. Otherwise, every enterprise would have removed all their desktop PCs by now, and they haven't. But desktop virtualisation could move further in that direction, saving energy and maintenance costs in the process. So why hasn't it?

My first conclusion is to agree that vendors need to explain it better. My second, more wide-ranging, perhaps self-serving but linked, conclusion is that vendors' marketing budgets should be directed more towards helping independent industry analysts and journalists, who can in turn help customers make the right choices for their own businesses.

Vendors and customers would ultimately benefit -- and, although ZDNet might be a beneficiary were it to come true, it doesn't invalidate the conclusion.