Symbian's Secret History: The battle for the company's soul
How Nokia took charge, and never let go
Part Two This is Part Two of a special report about the formation and early years of Symbian. It contains stories and plans that have never previously come to light. Part One dealt with the formation of the venture. Here’s the main course: how the decisions taken between 1999 and 2001 shaped not just the future of Symbian, but the course of the mobile industry over the next decade.
Let's rejoin the story on launch day.
On a brilliantly sunny June day in London, 130 Psion Software staff arrived at work and found they had a new employer, with an unfamiliar name. Most had been completely unaware of the development.
The day before, they had been working at a tiny company, trying to crack into new markets dominated by giants. Now they learned that the mobile industry's giants had entrusted their futures to them. And Epoc, the operating system that had been put on life support three months earlier, now looked to have a boundless future. They were now at the centre of the fastest growing, wealthiest sector of the technology industry.
It was the start of an intoxicating time. Two years later, Symbian would reach a valuation of £11bn, by some estimates, and be cited as the most important technology company on the planet. It seemed to be a new kind of business, and a validation of the European model of competition based on co-operation at deep technical levels, the model that had created the GSM mobile success story.
Symbian owned the future. Or so it seemed.
Yet Symbian would not ship a mass market product for four years. By then, many of the visionaries who founded the company had departed, and Symbian - while validating the original idea - had become very different to the original dream. Thanks to never-before-seen designs, and the recollections of many of the executives involved, this story for the first time explains how this happened.
The world reacts
Symbian sprang into life with two levels of management, representing operational management and owners. The operational board was led by CEO Colly Myers, with an all-Psion team of Stephen Randall, Juha Christensen, David Wood and Bill Batchelor, who had led the Series 5 project. Later, Thomas Chambers joined as CFO and Kent Eriksson came on board too. Psion CEO David Potter acted as non-executive chairman of the owner’s board.
So the Symbian management had an all-Psion operational board, but a Psion-free company board. It's not such an unusual arrangement. What made it unique is that the company board were all customers of the people they managed. This was to cause enormous tensions, as we'll see.
It meant that getting everybody rowing in the same direction was difficult. The customer-owners could collude to make decisions over the heads of the operational staff, or behind closed doors, and then enforce them from on high.
Still, the early days were promising. The dominoes just seemed to fall into place remarkably quickly. Former Symbian executives say they made a conscious decision to attract licensees, rather than ship product, to give the venture momentum – by stopping them signing with Microsoft. By October, Motorola had formally confirmed its membership. Japanese giant Matsushita joined in the following May. Symbian had a very public royalty structure - £5 for a smartphone, £10 for a communicator - and vowed to keep this for all comers, no matter how small or inexperienced.
Palm responded with a sniffy press release, describing it as a "concept without customers". Bill Gates fumed in private. Two memos (a year apart) that subsequently surfaced show that the Microsoft founder perceived Symbian as a personal snub - and issued a range of responses... some rational, some less so.
Psion’s share price rocketed, with analysts valuing Symbian at £4bn.
“The second tier companies assumed the first tier companies must have done their due diligence. Within a year we had 80 per cent of the market,” Randall recalls. That was the existing market – the new market, for which Symbian was created, didn’t yet exist.
The operational board of former Psion execs had a fairly clear idea of what Symbian needed to do: to achieve goals that the members couldn’t do individually, or together by other means. But beneath the surface, there were hints that the different shareholder companies had different ideas.
Symbian was set up to create something that didn’t exist: a mass market for smartphones and mobile data devices. These needed slick user interfaces and applications. For developers, the experience of coding for Epoc was a challenge. It was full of peculiarities and ingenious hacks – but developers used to slick Microsoft or Borland tools expected something better.
Symbian’s operational board identified a company that could help. A Canadian outfit Metrowerks had helped saved Apple in its transition to Power PC chips a few years before, and had created a developer friendly IDE called CodeWarrior.
But the Symbian executive board, comprised of the shareholders, vetoed the idea. Former executives say Nokia wielded the veto. Metrowerks was snapped up by Motorola.
The war over the UI
The user interface issue became the most hotly contested topic at Symbian. CEO Colly Myers said that successful Symbian devices must have the “enchantment” factor – explicitly recognising the importance of user experience, and high product design values. Twelve years on, this is universally recognized.
But it was harder to put into practice.
On Symbian’s launch day, Myers spoke of a desire to create several device “families”, and said there may be between “five or eight”. He stressed the new company's philosophy was flexible. Unlike Microsoft, Symbian would encourage licensees to innovate and customise on top of these. He was sincere – it was a cornerstone of the venture, and without the assurance, Symbian would never have coalesced.
There were practical reasons to keep the number of “device families” small. Developers prefer a small number of build targets for which to design their applications. Make too many, and the administration and support burden for a developer multiplies exponentially. Application developers can’t use a specific button or key shortcut, for example, if they're not sure it's going to be there.
Today, former VP of technology at Symbian Simon East is developing iPhone applications and finds it remarkable that Apple has just one build target for four generations of iPhones, iPod Touches, and even the iPad. But then Apple licenses its software to no one.
So how many "families" was too many?
Sponsored: DevOps and continuous delivery