I’m usually too far behind on the news to take part in fashionable discussions, but let’s try for a change…
The thing is, it seems I can’t open my RSS reader these days without finding yet another blog entry on comments made by the CEO of Red Hat, who was explaining why Red Hat will not engage in a battle for the desktop:
“First of all, I don’t know how to make money on it,” Whitehurst said. “Very few people are running a desktop that’s mission-critical,” so they do not want to pay the company for a desktop OS, he said.
Although some may disagree, claiming that Microsoft’s and Apple’s success in this space disprove these notions, I guess Whitehurst is mostly right about this. After all, Microsoft get paid for their OS because OEMs don’t give customers any choice (besides, when given choice, they resort to illegal copying) and Apple get paid for their OS because, well, Apple certainly won’t give customers a choice. But how many of these customers would throw in another hundred quid or so to get software support services with that? That is the crucial question, because, as several commenters point out, you’re not likely to sell many boxed copies of a free operating system.
Christopher Tozzi argues that IBM’s Ubuntu offerings suggest that there is indeed money in desktop Linux. We shouldn’t forget though, that IBM is ever so slightly bigger than Red Hat, at roughly 40 times the market cap., and can thus afford to experiment a bit more. Besides, they have a second agenda here (if not also a third and a fourth…): Ubuntu can help them break free from Microsoft’s chains where it comes to pushing Lotus Notes. So even if they burn some money up front, IBM will always gain something from investing in the Linux desktop, whereas for Red Hat the risks are relatively bigger.
But to be honest, this is not the interesting part of the discussion. Mark Shuttleworth also thinks he won’t be making money on the desktop, but that obviously doesn’t stop him from heavily investing in it.
On to other reasons then. Quoting from the Infoworld piece again:
[…] Whitehurst added he was uncertain how relevant the desktop itself will be in five years, with the advent of concepts such as cloud-based and smartphone computing and VDI (Virtual Desktop Infrastructure).
Also, the New York Times reported some further quotes of Whitehurst’s today (see, I’m not totally behind on the news!):
Mr. Whitehurst argues that Red Hat has shied away from products like netbooks because it doesn’t want to take a brand hit. […] “We have an extraordinary brand, and the last thing I want is for someone to take a computer home and find something is not working”
It all seems to make a lot of business sense – protect the value of your trademark, and plan for the future. Or does it?
As for protecting the Red Hat brand: I think that remark is merely an easy digg at the competition. In fact Red Hat found a solution to this objection a long time ago, when they created the Fedora name. They could invest as much effort as they want into furthering desktop Linux within the context of the Fedora project, without ever hurting their precious brand.
But it is on the other point that Whitehurst really errs: the future of the desktop. Ok, maybe the whole cloud thing (*cough*) is going to be the real money maker, and maybe the future won’t have enterprise desktops. But the thing is, the desktop is the key to that future:
Through the desktop, you achieve lock-in.
I’m not kidding. Everyone’s hopes are up that now that governments are all becoming open-standards aware, we will soon be rid of that nasty phenomenon of vendor lock-in, but I think document formats were never the problem: the users are. I might just elaborate on that in another post sometime, but for now, simply think about all those tech-support people you’ll need once the small-business world starts adopting open source software on a big scale.
Those demands won’t be met by the rather small supply of geeks who can find their way in every OS and every distro within days. So a services-based company, like Red Hat, or Canonical, needs to invest in creating a healthy employee market. In fact, once the adoption of open source really soars, they’ll need all the hands they can get – including many that only ever obtained their MSCE after working for it very hard.
This is where the real costs are: if learning to work with a system is somewhat of a struggle for you, then there’s a big personal economic cost that you incur when you have to switch. This is also a point that Microsoft and Apple understand very well: do you really think those massive discounts for the education sector are just a form charity? And why doesn’t Microsoft go after individual copyright infringers the way the RIAA does?
It’s a lock-in strategy. Not just user lock-in, mind you. Job market lock-in.
By the time kids start struggling for their MSCE, they actually already have years of Windows admin experience under their belt, because on the home desktop, everyone is their own admin. And despite that, many of them have to really work hard for it. I’m just saying, there’s no way they could all make a quick switch and be a Red Hat admin next year.
I guess you see where I’m going by now. Just to really hit the message home: do you think Edubuntu is just a form of charity?
So, Mr Whitehurst, ten years from now there may not be a single desktop left in the enterprise, but you might find that it’s more expensive than you thought to make your new employees un-learn their APT, Launchpad, deb-packaging, Bazaar, and Landscape skills. To tear them from their established social context at Ubuntu Forums, and to rewrite their by then almost hard-coded array of resources (debian.org, backports.org, etc etc).
I may be wrong though. I mean, I don’t know *** about business.
But it was fun writing this pretending I do, anyway :)