25. The merger that failed
The “Sphinx” project to release v9.5 of Symbian OS vividly illustrated the problem that a merger of the development teams of Symbian and Nokia S60 was intended to solve.
The Sphinx project completed at the end of September 2008, apparently in great shape. As covered at the end of the chapter “Google on the horizon”, Sphinx delivered its agreed scope, within the allocated time and finance budgets, and with all its quality targets being either met or exceeded.
Nevertheless, the development teams in Nokia decided not to use this version of Symbian OS. As Per Selbekk, Symbian’s EVP of Software Engineering reported a few days later:
As you might have heard already, Nokia has made the decision not to use Sphinx for their upcoming Timebox 9.1 release. We have worked hard since the summer to see if we could get Nokia to use Sphinx instead of Naiad in TB 9.1 but, in the end, this was not possible.
I would like to make it very clear that Nokia’s decision is not a reflection of Sphinx’s quality or the work we have done to help Nokia use Sphinx. On the contrary, Nokia has been very pleased with this release’s quality and the assistance given, specifically on 3-plane comms which we have got up and running on several hardware platforms.
In the end this was a programme decision and not one based on technology. Nokia has certain handset milestones to reach based on TB 9.1 that could not be moved and internal resource constraints meant that they would have to stick to Naiad [Symbian OS v9.4]…
The work to create Sphinx wasn’t altogether wasted; much of it would subsequently find its way into later Nokia “Time Box” (“TB”) releases of the S60 software. However, seen from the point of “lean delivery”, it was far from being the best use of Symbian resources:
- The software in Sphinx was delivered earlier than it could be incorporated in products being released to the market
- As such, Sphinx components became “work-in-progress inventory”: items waiting to be picked up and used, but which are subject to a lessening of their value as time passes and as customer and market requirements change
- In due course, when these components are integrated into product ready to ship, issues are likely to arise which are no longer easy to resolve, because the original developers are no longer available to work on updates
- Far better, in this view, for the Symbian developers to have worked instead on software that would have reached the market sooner.
In other words, it would have been better to plan and design Symbian OS components as part of a larger, integrated planning and design process that included S60 components too. Rather than focusing on Symbian OS in its own right, it would be better to consider enhancements to what was now called “the Symbian platform”, which included S60 as well as Symbian OS.
When Symbian had significant sales from all three of its live UI systems – S60, UIQ, and MOAP(S) (the latter being used in Japan), it made more sense to design the OS layer separately. But recognising the market reality that S60 was now by far the most important UI system, it was time to set aside this separation.
For that reason, managers in Symbian eagerly looked ahead to more productive working relations with their new colleagues in Nokia’s Devices R&D department, as the “Seaside” merger became a reality. There was a great deal that the two teams could expect to work on together:
- Accelerating support for the Symbian platform on hardware from both Qualcomm and ST Ericsson; Nokia had ambitious plans for sets of new breakthrough Symbian devices incorporating chipsets from these manufacturers
- Improvements to the user experience, including a “Direct UI” touch-based interface in which more actions happened with a single touch (rather than requiring two touches)
- Improvements to the ease of developing applications, via the Qt libraries from Nokia’s acquisition of Trolltech
- Re-writing, modernisation, and in some cases replacement, of many of the intermediate components of software that had been created several years earlier, often without much regard for longevity or scalability
- Two-way sharing of “best practice” development methods, with each of the two sides to this merger adopting techniques and processes that had been proven to work in the other side.
The fact that so little of this happened as expected is a sad example of the general principle that mergers often turn out to be more difficult than expected. This chapter looks at what happened.
Nokia’s “Seaside” initiative to unify the software teams from Symbian and S60 was only one of several initiatives they were undertaking, throughout 2008 and 2009, to improve their software development capability. The over-arching improvement initiative was named “Ocean”, run directly by Peter Ropke, Nokia’s SVP for Devices R&D.
The intent of Ocean was to ensure that Nokia was best equipped to operate in the fast-evolving competitive landscape that was anticipated to unfold over the next few years. The team took 2012 as the end point for their vision.
As part of Ocean planning, Peter Ropke organised a two day “transformation workshop” in Helinksi’s Hilton Kalastajatorppa hotel, on 8-9 December 2008, and invited around 40 of the senior managers from his own extended leadership team and from Symbian.
[ SNIP ]