As if that wasn't bad enough, last week the USB-attached hard drive that I've been using for important backups started failing. It seemd to be suffering from some sort of corruption in the boot sector. Luckily, after enduring four or five days of pain this week slowly retrieving the data from the misbehaving drive, I've managed to recover virtually all of the data. (Tips about what I've learned about disk recovery will follow in a forthcoming iTWire article.)

One of the buzzwords in IT these days is "cloud computing" and there's much speculation as to whether it will (some time or other) take over from the Microsoft-style desktop personal-computer-based computing scenario that began around thirty years ago. I've just been reading How Cloud Computing Is Changing the World in BusinessWeek:

"At first, just a handful of employees at Sanmina-SCI began using Google Apps for tasks like e-mail, document creation, and appointment scheduling. Now, just six months later, almost 1,000 employees of the electronics manufacturing company go online to use Google Apps in place of the comparable Microsoft tools."

This adopt-by-stealth approach started me thinking about this topic in terms of the "good enough" theme of this one of my blogs. Should we seriously be considering a widespread move to this way of doing things?

It may well be the way that many more of us gradually adopt this means of computing. At least, we  for the 80 percent of situations when we find the facilities being offered to indeed be "good enough" for us to accomplish our computing goals. Desktop computing may well then be relegated to the 20 percent rump of specialist computing needs such as intense graphics or intricate software development (but maybe even most of these might be offered in "the cloud" by niche cloud computing service providers).

It's not just a matter of the application software features being adequate. The underlying infrastructure will need to improve to the point where we're totally satisfied that it is "good enough" for our purposes.

This encompasses not only such things like speed and response times of the base application service itself but also the supporting aspects: network capabilities (presumably broadband of some sort); service level guarantees (a Google-like network of highly-redundant servers and networking pathways, guarantees for reliability and fast restoration of data after a crash or human error); and most importantly ultra-high levels of security and privacy if the user expects it -- but probably isn't such as issue in social computing applications, where openness is the norm.

Why outsource your critical apps to a cloud provider unless you get the level of service that you expect?

PLEASE READ ON...