Nothing this year then:
http://www.nats.co.uk/news/releases/2005/2005_02_16a.html
Lets just get this right. In BW view of systems development:
1) Projects should overrun by years, because they all do.
2) Projects should overspend by at least 30% because they always do.
3) Everything can be fixed in later bug-fixes.
Quick frankly that is absolute cr*p. That is what cowboy plumbers and builders do. Oh don't bother testing it, it will be tested when they use it. Don't bother prototyping the screens with real system users, the managers that never use the thing can sign it off against a paper specification and the users will like it.
I am a Solution Architect (previously Lead Systems Developer) working for one of the largest computer services companies in the world. We work with the largest companies and governments and we are heavily commited to a concept that is apparently absolute heresy to you:
Right First Time. To this end we achieved SEI CMM Level 5 back in 2002 (we have employees seconded full-time to the SEI), but even without such procedures it is possible to develop systems that do not need "gradual improvement" to resolve faults.
A decent system should not need ANY ongoing support effort, only enhancement for new functionality. From personal experience I have had 3 faults reported for software developed by myself over the last 4 years, total, across 7 systems being used by 10,000 users at 4 different multinational customers. And none of those reports were due to faults in the delivered software, but were consequences of the BW style of working by others at the customer sites. Server names were changed or user ids wiped out without prior notification to potentially impacted system owners - they just assumed that "if anybody was affected by the change then they would complain about it", without thinking about the total cost to the business of their incompetance.
Get your head out of the sand: you obviously are totally unaware of what the word Quality means or that the concept even can be applied to system development. Yes that is what happened in the 70's, but 30 years later it is perfectly possible to deliver a fully tested, fault free system first time, and once you understand how to do it, the systems are delivered earlier and at less cost. Yes, systems delivered early, fault free and at less cost than the detailed plans said it would be. It happens, and what happened at Swanwick was incompetance pure and simple