The news last week that IBM had challenged its internal tech team to roll out desktop Linux throughout the organisation by the end of 2005 seemed like a real shot in the arm for the whole open-source-client alternative to Microsoft -- right up to the point when IBM began back-tracking faster than Tony Blair discussing Ken Livingstone.
A November memo from CIO Bob Greenberg, leaked to the Inquirer news site, reportedly said IBM chairman Sam Palmisano had "challenged the IT organisation, and indeed all of IBM, to move to a Linux-based desktop before the end of 2005." But in a furious bit of spin management, IBM spokeswoman Trink Guarino jumped all over the story and said that IBM was merely evaluating the idea of desktop Linux. "IBM has no plans to move all of its employees to Linux desktops by 2005," or even a majority of them, she said.
So just when it seemed that someone was prepared to put their head above the parapet and give the desktop Linux movement a hefty leg-up along comes the PR machine and yanks them down safely behind the battlements again. That's really inspiring for anyone thinking of trialling the technology, and surely a marketing triumph for Big Blue.
There must have been a furious exec meeting after the news of the leak broke, where IBM had a choice of three paths. One: stand by the document and its gutsy message; Two: fudge it and be non-committal; and Three: deny everything. Well, Big Blue settled on the middle road and showed once again why Microsoft has managed to dominate the desktop for so long. IBM UK wouldn't even comment on the story. Impressive.
But what about the rest of the big vendors? Well, Novell has made some of the right noises but hasn't as yet summoned up the courage to commit to a time-frame for a wholesale internal adoption. Speaking with Novell's UK managing director this week, I heard that the UK organisation was "aggressive" about adopting Linux, and welcomed the chance not to have to pay Microsoft any more licence fees, but wouldn't be drawn on when this might happen.
Which is kind of frustrating, because if anyone has got the motivation and opportunity to make the grand gesture to the rest of the industry and start eating its own dog food then it's Novell. This is a company that bought Linux tool vendor Ximian, followed by its very own Linux distributor in the shape of SuSE, and which has a very turbulent history with Microsoft. Novell should relish the opportunity to hit the Redmond boys where it hurts and be seen as a real gutsy innovator at the same time. The best I could get back from the MD was to wait and see what turns up at BrainShare in March when the company traditionally makes most of its big announcements.
Talkback
Its disgusting to see these spineless CEOs announcing savings of using O.S. and even thumping chest on how much money they made (HP, IBM) ! See the paradox there ?
IBM has always been FUDing to O.S. community - they know that most open source/linux users use Unix at work and they want to replace any brand of unix ( even Aix) with their 'linux' ( read proprietary) solutions, usually costing MORE than what they are trying to replace. Each unix installation lost is a gain for m$. For m$ will later market the hell out of linux and no vendor will be there to defend the use of linux.
Take the example of Munich - they (IBM) charged MORE than m$ for the contract. Now they are not even offering support to migrate ( m$ deal appears to be better actually) ! Pretty soon munich will return to m$ ( there are already reports of bickerings), IBM would have made its illicit millions and linux advocates will still be clinging to IBM coattails for an encore.
IBM DOES NOT advocate linux, it takes ADVANTAGE of die hard linux advocates, including city of Munich. If IBM was interested in advocating linux, it would have used this opportunity to do a good job at cost savings to munich; instead they actually seem to be helping m$ by a) charging higher than m$ and b) screwing it up.
IBM 'lead' the personal computer market AFTER seeing the puny Apple making money 'it should have'.
I trust that one of the 'big guys' - IBM, Novell, Dell, or Gateway, et al., - will 'lead' the market AFTER one of them sees a tiny upstart 'rake in the dough' from their linux efforts.
The problem with the Gartner figures regarding Linux on the desktop is that they have applied the use of Linux under a traditional Windows desktop deployment model. Of course the traditional model of Windows desktop deployment (with acutal software installed on each computer) is expensive to run whether it's Windows or Linux and fails to take into account the powerful elements of the Linux environment.
For example, simply switching to a Terminal-Server model of deployment and significantly cutting down on the number of desktops that must have software maintained on them would offer consisderable savings for starters. Deploying this model is simple with prerolled districutions like LTSP (www.ltsp.org) already available. Even better, Windows applications that can't easily be replaced can be served up from a dedicated Windows server allowing access to these applications from a Linux desktop.
While Gartner is keen to look authorative in their report, the numbers only add up if you stick to the traditional Windows desktop model. Linux doesn't just offer the opportunity to use suitable cheaper software, it also offers the opportunity to reconsider how you deploy you desktops and offers flexibility along with considerable savings.
PS. Windows Client Server and Linux Client Servers can't really be compared. Windows CS models have so many limitations that they are unweiled to deploy, while Linux CS offers a flexibility unrealized by many Windows sys-admins.
Baby steps, man.
Including the aforementioned vendors, how does any networked company roll out what is essentially a paradigm shift? Or commit to an arbitrary date for migrating business processes to new systems for employees already preoccupied with keeping the wheels of commerce at the company turning? It's naïve to suggest it is even possible to move any company's diverse enterprise network user audience lock stock and barrel into the open source world. Even if everything worked technically, there would be unacceptable business interruption because Linux is completely new and unfamiliar to 92% of all computer users. Most Windows-using people wouldn't do well with "new" technologies when it comes to doing their jobs because they've learned Windows. That's why OpenOffice is a good place to start conditioning garden-variety Windows users to the GNU world.
Besides the business hiccup, what organization is prepared to take the bullet for replacing current desktop technologies-investments often in the tens of millions--to further accelerate what is largely a fait accompli? Clearly Windows is insecure, proprietary and legacy, but when it's patched and stable it works and in most companies it's paid for. Open source has a definite role emerging on the enterprise desktop, and that need will be filled incrementally by design. But they probably won't be replacing Windows desktops en masse until PC lifecycles are exhausted or Windows OS upgrades are required. And only when the company is ready to support or pay for replacing the desktop hardware, software and key apps with stable, rich Linux equivalents and then commit to retraining specific groups of users.
Deploying Linux broadly on the desktop isn't nearly as important as it might appear, nor should it be. Both vendors and customers have a far greater near-term opportunity to deploy server-based applications on Linux and reap the benefits with little to no business interruption. After all HP did 2.5 billion last year, and very little of that revenue was on the desktop. For CTO bean counters, that would be known as the lowest hanging fruit. For journalists, it's yesterday's news. But it bears repeating, especially now that Linux has a bottom-to-top offering and can-do vendors behind it.
Let's start at the bottom of the stack and build a better, less predacious networked world from the ground up the next time around, okay?