Yet another plan to revolutionise investment in the railways was revealed this month - this latest from the Strategic Rail Authority.
The big idea in the SRA's Strategic Agenda involves limiting Railtrack to managing maintenance and renewal of the network, and handing major upgrade projects over to privately financed 'special purpose vehicles' to build and then sell to Railtrack to operate.
On the surface this sounds like common sense. A cashstrapped Railtrack would not have to raise the investment money. Nor would it take any construction risk so would be able to borrow the cash to buy these finished projects at lower interest rates.
But scratch this surface and a heap of complications is revealed.
When most people think of major rail projects they think about the Channel Tunnel Rail Link - effectively the only model for the SPV idea in the rail business. A private sector consortium is building the railway with Railtrack as the ultimate purchaser.
But it is worth remembering that CTRL had to be bailed out by the government and by Railtrack a couple of years ago when the revenue and cost projections failed to stack up.
However, CTRL is not really representative of typical UK rail upgrade work. CTRL section one is a brand new, largely greenfield railway. How many of those are we likely to see in the next 15 or 20 years?
In reality, upgrading Britain's rail routes will mean linking up many smaller projects, each carried out on an operating railway - a revised junction layout here, a bit of track straightening there.
And as more train services demand more capacity the down time for possessions is increasingly limited. Any possession is used at the moment by Railtrack to the maximum, to maintain, renew and carry out major project work (see feature p31). The tasks are inextricably linked and cannot conveniently or sensibly be separated out to suit a new investment model.
Nor can procurement. The SRA assumes that the constraint to redevelopment of Britain's railways is money and project management skill. But ask engineers at Railtrack what is holding them back and, after the lack of possession time, they talk about the capacity of the rail infrastructure industry.
Railtrack and its program managers - the companies brought in to address its acknowledged management skill shortage - are working with suppliers to get the maximum out of limited resources by boosting efficiency. And although Railtrack has also been looking to bring more rail contractors and designers from overseas into the UK market, there are simply not enough of them.
So any idea the SRA might have of throwing a lot of individual investment projects with myriad different clients out to separate tender would appear to be a recipe for chaos and price escalation.
The SRA needs at least to talk the detail of its proposals through with Railtrack's engineers and listen to their advice before it proceeds. Railtrack wants to bring SPVs into the mix but the SRA must accept that Railtrack engineers have the experience of the issues involved in doing that.
As it stands, the neat and tidy split in workload proposed in the SRA Strategic Agenda is unlikely to be the most efficient way of forward.