Not having studied psychology, I'm usure about the correct terminology but it's very interesting to me the notions that some people have about uptime at their organisations and the knots they tie themselves up in to jusitify keeping systems in house.

Quite often -- especially in discussions about cloud computing -- I hear comments made about system availability that are basically just fantasy. They say 'My CEO would kill me if I had any email downtime'. And yet just the other day the same company couldn't send me email for a couple of hours.

Maybe they're using words to disguise uncomfortable realities so, for example, it might be literally true that there was no problem with the email server itself, just the network. Or that the downtime was planned downtime so (at least internally) people could work around a known issue.

And then when it comes to comparing against hosted services, there's often a level of inventiveness that comes in. A CIO recently told me that Google had suffered huge outages for its Apps service that knocked users off mail for several days, for example. And yet, given the viral nature of social networks, it is nigh on impossible for this to have been true and the 'fact' to go unreported (or at least un-Tweeted).

The interesting thing is that I don't think these people are being consciously economical with the truth; they are just very afraid of losing control of a chunk of infrastructure and jumping at factoids to suit dearly-held tenets. This is understandable: as well as losing a fiefdom, they might well be having to say goodbye to friends and colleagues as part of a decision to outsource what has quickly become a commodity service.

The move to services like Google Apps,, Windows Live and so on is, of course, a cultural change as much as anything else. It changes the way companies pay for IT, manage it, hire around it and so on. So it's not surprising that for many it will be an emotional decision that brings in feared change processes. The problem is that, in areas such as email and CRM, that change now seems inevitable and those that put off today are merely delaying the inevitable.

Of course, some fears are grounded. Internet access is not literally ubiquitous and levels of reliability and availability will vary. This is why vendors are adding granularity (sorry, terrible word) to their services so that if there are issues they can be worked around (via offline components, for example) or isolated.

Suppliers can also help CIOs work around these fears through transparency. Using Salesforce as an example, they need to openly document downtime, geographies affected and so on. As much as anything their role will be to become assistants in the change process from client/server to internet-centric computing.