Mythically, there’s only been the one British public sector procurement project that arrived under time and under budget - Polaris. The myth is around the “one” for historically there may well have been others. But in modern times it probably is true.
The secret in that project was the importance of interfaces. Any part of the system could change their design, build, method, whenever and however they wanted. But changing the interface with any other part of the system was between verboeten and extraordinarily difficult. Everyone could - and did - do what they thought best, the limitation on changes was when such a change would act upon those similar freedoms of another.
The allegory with the free market, the liberal society, is obvious. Do as you wish except when it affects others. Supply goods and services by whatever method, it’s that interface with the wider population, the market, that matters.
But more narrowly, it’s also the thing that British government planning doesn’t do these days:
The UK’s beleaguered public libraries have been let down by years of indecision and delays over how to spend millions of pounds in funding earmarked for a nationwide website.
This was among damning criticisms voiced on Saturday by campaigners who have lost patience with the government, the British Library and Arts Council England (ACE) over their longstanding failure to develop a nationwide scheme. The “Single Digital Presence” (SDP) – renamed LibraryOn – was meant to bring together public libraries in one website to enable the public to access collections across the country.
The problem has been that there are 150 library authorities in England alone, each with their own technology and management systems.
The problem is that everyone’s been arguing about what such a site should do. Which goals are to be achieved? Rather than concentrating upon how to do it, which is a task of the utmost triviality.
All of the libraries are digitised - in their records at least. All have booking systems. All have the ability to create a queue, to allocate a newly freed up book to the next in it and so on. So, all of the hard work is done. All that is required is to enable all the 150 systems to speak to each other - the interface.
Simply define that interface. A library system should be able to output book data in this format here. Every such system should be able to import data in that same format. There, we’re done. This is something that could be knocked up over a weekend by the readership of The Register.
To emphasise the “another” in that headline. Back when there was that £11 billion to spend upon digitising the NHS. That £11 billion that was spent without delivering one single usable line of code. The mistake was to try to design that system. Instead of simply publishing the interface. All medical records systems must be able to out to this format, all must be able to import from it. There, we’re done. Over time and over the normal replacement and capital cycle all machines would then talk to each other across the NHS and we would have, successfully, digitised medical records. There was even such an American standard - a free one - that we could have adopted.
We are, just occasionally, willing to admit that there might be a purpose, a use, for government. Say, in the creation of interfaces which then allow everyone to get on with their own thing. But don’t, as experience keeps insisting, allow government to ever actually try to design or plan anything.