There are very few things bloggers or analysts can do to erode their credibility with experienced IT managers than to say some new technology is going to change every aspect of computing within a very short time.

IT people have just heard that way too often and know that even when it's true, the change isn't always an improvement.

So I won't say it (though I may imply it heavily.)

It's a toss-up right now whether the techniques of virtualization are squeezing into more and more technology areas, or whether more vendors are just identifying the things they've been doing all along as "virtualization."

Case in point is this guy, whose SOA conference presentation is titled Web Service Virtualization for SOA Runtime Governance and Control. Nice tie-in to technology of the moment, and there may actually be an underlying connection (though, granted he's a vendor and may therefore be more driven by marketing than reality).

I thought the whole point of SOA and Web Services was to abstract the function of the software from its APIs, platforms, location and other limiting factors.

Virtualization was the whole point of the World Wide Web, in fact, which has found some moderate success abstracting the location, native formats and access rules limiting access to information and making it available to anyone who can click a mouse or operate a search engine.

Virtualization was the whole point of middleware as well, which abstracted hardware from software, speaking to each in its own native language so neither had to do it themselves.

Ditto for the other common networking and software protocols; SMTP, IP, Ethernet, distributed directories, the Domain Name System, storage area networking SNMP, POP3, TCP, URI, OLE, SOAP, blahblahblah (and BlaBla2.0, of course).

Half the new technologies in computing over the last 40 years have been efforts to make connections across gaps created by the other half. Most of them have been efforts to virtualize the connection between one thing and another by adding a layer or language that made the gap irrelevant.

So I'm not uncomfortable predicting that virtualization will squeeze into every corner of IT—because it makes a lot of things a lot easier.

But it is going to change things. It's going to change a lot of the decisions you make by eliminating the PITA factor. If you can do something well with great difficulty, or do it adequately with less difficulty, which are you going to do?

That sounds like criticism of the energy and determination of IT people, by the way, but it's not. If you're responsible for someone else's money—which, no matter how big, is what an IT budget amounts to— and it's possible to get 80 percent of the function for 20 percent the cost, it's irresponsible to go for the more expensive alternative without damn good reason.