Re: This lot almost make m$ look good
And that's why "big bang" implementations on this kind of scale are stupid and should never, ever be attempted under any circumstances whatsoever*.
You only find out that it's an abject failure once all the money has been spent with nothing to show for it, and either the customer pays huge cash for nothing or the supplier goes bankrupt.
Instead, if you do the project as a series of small changes, implementing each part of the system in a limited area (not the whole country) and expand it slowly, not only do you get a system that works (and is useful at each step) but there are many more companies that could bid for it because there's money at each stage so crazy credit isn't required.
As a bonus, when the credit requirements are smaller at shorter periods it's much cheaper to build, with less risk to all parties, and more of the cash goes towards the actual project rather than the banks that loaned the money.
As a bonus to the bonus, you can pause or even stop the project at almost any stage and still have something useful, and usually change the requirements of the next stage without penalties in time or money.
Yes, I just more or less described the Agile methodology.
It does work, unlike Waterfall which always fails.
*(C) Department of Redundancy Dept.