Bristol City Council’s failure to deliver on its open source strategy is beginning to make the coalition government’s manifesto commitment on open source look incontinent.
The council’s own open source strategy is looking ineffectual. Bristol Council cabinet committed to an open source infrastructure a year ago – as long as it was doable. It ordered a pilot but that was discredited by an allegation that it had been fixed. Now the council has refused to release the suspect pilot reports under Freedom of Information, it is time to look at those allegations in full.
Mark Taylor, CEO of Sirius, told MPs in May how, left to establishment suppliers Capgemini and Computacenter, the open source strategy got caught in a thicket of indifference and vested interests.
Bristol set an original deadline for its open source strategy to be costed and risked by November 2010. They told Capgemini to get on with it, said Taylor, but Capgemini did nothing. Bristol told Capg to work with Sirius, who were experienced implementing open source infrastructures for companies like SpecSavers. They ignored Surius, said Taylor in a letter to MPs on the Parliamentary Administration Select Committee (PASC).
The council meanwhile tendered for an open source infrastructure using the £6bn Buying Solutions Framework for Commodity IT Hardware and Software (CHITS). Thus it would be ready to roll when the pilot produced its recommendations.
Five large CHITS suppliers bid for the work: Computacenter, Fujitsu Services, Insight Direct, Softcat and SCC. Sirius complained it was forbidden from bidding for the work because it wasn’t on CHITS, a list dominated by the UK’s largest infrastructure suppliers. Sirius teamed up as Computacenter’s open source advisors and the pair won the business. But things then took a turn for the worst, said Taylor’s letter.
“ComputerCenter (sic) began the project by having a series of meetings with Bristol City Council without inviting Sirius or even mentioning that the meetings had occurred.
“Sirius first met with the Council on 17th January at a meeting billed by ComputerCenter as the ‘kick off’. On the same day we were included on an email from ComputerCenter’s Project Manager referring to earlier meetings and assuming all recommendations to the Council would be proprietary software as usual,” said the letter.
From the very start of the work in January, Sirius was already writing formal letters of complaint to Bristol over Computacenter’s bias for the proprietary software vendors with which it had a business relationship and against open source.
This bias was a hot topic in Whitehall. Within a month, the Cabinet Office called the UK’s systems integrators in for a telling off: why had they been ignoring government requests for open source systems for two years?
Capgemini had not even started the pilot, said Taylor’s letter. It had been five months.
So Bristol asked Computacenter and Sirius to do the pilot themselves before implementing its findings. What transpired so frustrated Taylor that his protests put him at loggerheads with Computacenter and had him kicked off the programme.
Taylor complained that Computacenter – one of Microsoft’s largest UK resellers – had skewed the pilot’s parameters so it came out in favour of Microsoft. Implicit in this allegation was the assumption that Computacenter’s profit margins relied on the supply and servicing of proprietary software systems, that these systems raise a relatively high income from their monopoly rents and that an industry established on this model had no interest in alternatives.
It went right to the heart of the commitment written into the government’s coalition agreement to “create a level playing field for open source”. The industry was founded by companies like Computacenter on proprietary software by publishers like Microsoft and its business model had become thus fossilised. The open source model was disruptive. The proprietary software industry wasn’t going to let it in, no matter how discredited it had become for its cost, waste and failure.
“ComputerCenter showed a clear and persistent bias towards proprietary (invariably Microsoft) software,” said Taylor’s letter to MPs.
ComputaCenter turned the pilot into “an enormous and costly paper-based check-box exercise” that dragged on for months longer, he said.
But here was the key allegation of bias: Taylor said Computacenter had made the “starting point” of its pilot, “a comparison between how well a Microsoft software stack and an Open Source software stack compared with a Microsoft software feature set”. On those terms, nothing could win but Microsoft, not even the largely Novell-based infrastructure the council was already using.
Computacenter in addition “lobbied Bristol to buy a proprietary, not an open source infrastructure. Taylor said Computacenter made him sign a contract forbidding him direct contact with Bristol, the council that had turned his 2010 consulting advice into an ICT strategy championed in person by council leader Barbara Janke.
He was excluded from meetings about the pilot and not informed of others. He relied on Councillor Mark Wright, who authored the council’s ICT strategy, to inform him when meetings were happening. He fought against Computacenter’s insistence that the council should use proprietary software instead of open source alternatives.
In May, Computacenter submitted a recommendation that, said Taylor, the council use “an essentially entirely proprietary Microsoft stack”. Taylor complained and his partnership with Computacenter collapsed.
Computacenter claims the allegations are “potentially libellous” and refuses to discuss them, perhaps highlighting one of the most important aspects of this episode – the question of the government’s transparency agenda, as discussed elsewhere.
But the episode raises concerns also about the government’s SME-led innovation policy, something else championed by Bristol’s “Digital City” strategy. Ultimately, both issues are indicators of the state of the industry, fed as it is by school leavers whose computer education consists of learning to operate Microsoft applications.
As Microsoft celebrates the 30th anniversary of the IBM PC on which it piggy-backed to power it is worth reflecting on how it was, famously, the IBM platform’s openness that made it such a success. If anyone cares also to reflect on how the 30-year old industry might benefit from reform, they might find a lack of transparency will work to the detriment of them and everyone else.