I am doing some research on a big contract the British government-associated educational consultancy Becta will be awarding in the fall, some details of which were announced July 7. The press release and backup documents said the procurement would:

• Top $100 million in spending (at 2008 exchange rates) over four years. There are apparently some appropriate "economic/financial capacity" criteria in the bid qualification process that preclude smaller organizations not able to deliver at that scale.
• Include "both proprietary and open source software." That criterion supports the true open choice needed going forward but it is not a requirement that a supplier offer both, although it is to a supplier's advantage if it does.
• "... include(s) operating systems, network software, security and maintenance and enterprise resource software," not educational software. In other words, the software being procured seems to involve the complete range of "technical capacity" that every enterprise needs.

Based on this breadth and depth, I thought Becta might introduce some new ways to acquire enterprise software (and I still hope it may). Any new approach would be meaningful for all large dispersed IT organizations in the 21^st century. But my preliminary findings are that the Becta bid will result in an agreement similar to this one which was negotiated by the UK government for ERP, apparently for large organizations, and another similar agreement for enterprise software such as Becta wants for smaller departments and agencies.

If so, the UK approach is to acquire 21st century enterprise software the way software was acquired in the middle of the 20^th century: Buy it from services and systems suppliers rather than software suppliers. I would rather see software acquired using a new approach: Make it as easy as acquiring electricity. (And some enterprising Becta bidder might still come up with such a proposal.)

You don't want to own your own dynamo. Depending on your size, you might have to have your own electrician on staff, but that's not ideal either. You don't even want to put up with the software equivalent of 15-, 20- and 30-amp circuits (or whatever the equivalent is outside the U.S.). That is, you want to do away with the typical need to use different technologies for personal-productivity, back-office and front/extra-office enterprise software.

You can call it "cloud", but 40 years later, I still like the term Utility Computing that some MIT guys came up with. In the 21^st century, you want it all a plug-in away.