Saturday, August 11, 2007

Ruling on Novell's Unix assets bolsters OSS, binds Novell and IBM

So Novell really does now finally seem to own the Unix copyrights. Linux finds itself on a high-ground pedestal of long-term, low-risk use (unless Microsoft buys Novell [should have when they could have, eh?]). And IBM and Novell are closer than ever.

Fun times.

The folly of the SCO Group FUD fiasco hurt more than SCO's 2003 investors. It has shown that the PR war approach to undermining confidence in Linux and OSS is a loser's sport for wannabe spoilers.

And while Microsoft thought it bought Novell a pair of permanent knee pads with its Windows-SUSE Linux indemnification pact last year, IBM will now come around to stand Novell back up on its feet, perhaps for good.

Someone should write a novel about Novell's travails with Microsoft over the years. More plots than a first-year Fortran class.

And there's irony, too. IBM and Novell, for example. Who would have predicted 10 years ago that IBM and Novell/Unix/Linux would make great partners against Microsoft, Red Hat, Sun, Oracle and BEA? It's too good not to be true.

We just saw some first hard evidence of the deepening relationship with the IBM-Novell middleware and OS stack bundle news last week. And it's not too whacky to consider Novell's long, tortured journey finding a quiet resting place on the outskirts of Armonk, NY. Set the rocking chair with a nice view of the Hudson. Tea at 4 p.m., lovely sunsets.

I believe the slow-motion tennis ball lob on Linux risk FUD never quite made it over the net. The federal judge on the Unix case just knocked it back in Microsoft's court. SCO as proxy is no more. OSS and IBM's lawyers are the only winners. So let's close the book on that sordid chapter.

Much more interesting now is the market dynamics over how far and how fast up the datacenter stack open source components/solutions will move. Will the Red Hat/JBoss/Apache incubator code move the bar on OSS disruption into the echelons of serious enterprise middleware and beyond? Will data services get an open source foundation? Maybe an OSS metadata layer makes sense, just like OSS ESBs do?

IBM is the arbiter to watch on this transition. Novell can help them manage the timing, while covering a flank against Red Hat. BEA can only watch and wait. Oracle can time the infrastructure transition while nimbly moving up the stack to vertical business applications and data services. Ditto SAP. Sun can play around with trial-and-error open source models roulette while laying off its way to a niche hardware business.

HP could be very interesting. This will be the year for HP's SOA play. It needs to find a way to master the OSS/support/hardware/solutions/consolidation process. But where to chase for the next margins? Who is friend or foe? There won't be too much room for error on this one, not too many chances to recover from missed opportunities or misplaced bets.

The timing is key. And managing transitions from commercial to OSS up and around the stack (to ding competitors while remaining key to major accounts) is the game. There won't be any more Red Hat/JBosses, or such accidental empires. But there may be an OSS applications and services ecology on the horizon. And SOA will soon drive they types of choices that require businesses to focus on such an OSS services ecology.

So like the days when Unix was the infrastructure law in the core corporate datacenter (and Windows was only hype-ware there), we may be back to a period where the major transitions have little to do with Microsoft's rate cards. Microsoft will be at an ongoing disadvantage in the commercial-OSS transitional disruption march across back-end servers as long as it has no OSS strategy (other than FUD). And that FUD strategy has just come up wanting.

No comments:

Post a Comment