As Sun unravels, it's time for a new suit
Update: Sun says not so
Opinion With IBM reportedly breaking off the engagement, Sun must reinvent itself if it is to go it alone. That means bringing in a Joe Tucci lookalike to reform and restructure the business.
This is a highly personal opinion about how and why it should undertake that process. First of all, Sun's board and senior management need to recognise a few things that have become blindingly obvious to a lot of outside observers.
First of all, Sun cannot meaningfully monetise its massive developer download base. Those tens of millions of copies of Solaris, Java and other pieces of Sun software are bringing in diddly-squat revenues set against Sun's overall business and what it should be capable of doing.
Next, Sun cannot profitably survive in any sizeable way as a me-too x86 server vendor. It cannot prosper in any meaningfully-sized way as an open source software supplier and services vendor either, despite MySQL, Java and Open Solaris.
The Sun AMD play has run out of steam. Sun's hybrid server/storage play - the X4500 and similar products - is a limited niche offering that won't go very far. Sun's proprietary SPARC microprocessor architecture and the hardware built using it is severely threatened by Nehalem. Sun's mainframe tape storage business is a niche and always will be. Its mainstream disk drive arrays are me-too and going nowhere.
Their development has been crippled by Sun's focus on its Open Storage systems. This Open Storage effort, despite its coming in layered or tiered flash cache editions, with system software tweaked to use it, will not revolutionise the storage industry. Sun is wrong when it asserts that proprietary storage architectures will go the same way as proprietary server architectures and be washed away by commodity hardware and open source software, like Sun's Open Storage 7000 line. They won't.
Ditto Open Networking.
So, assuming these points are recognised, where does Sun go from here if it is not bought?
Continuing with its present strategy, which has been discredited by the company sale stories and by comprehensive business under-performance, should not be an option. Continuing with the present senior management team is also not viable. The first order of business by the board should be to choose a new CEO.
This CEO should have a deep and comprehensive understanding of the existing server, storage and software industry and markets, and be young enough to lead the company for, say, ten years of energetic activity. He or she should have rock-solid technology business leadership and execution experience, and a proven ability to take a ramshackle collection of underperforming businesses and turn them into a coherent and sustainably profitable whole.