On the heels of some of the recent Vista/Office 12 news, that’s really the question of the day, isn’t it? While few people are asking that directly, it’s the proverbial elephant in the room for most of the discussions that I’ve seen. As I see it, there are two basic camps here: Group A, that believes the desktop is poised for a revitalization through an Avalon-style blurring of the boundaries between local clients and network resources, and Group B that contends that Google is the way of the future, and the world is going to be one big network computer. As is my custom, I’ll group myself with neither. Oh, it’s true that I have a definite bias towards the thin-client end of the spectrum (I was an SI for too long, what can I tell you?), but I continue to be vexed by thin client shortcomings that convince me that I won’t be ceding all my data to the network cloud anytime soon.
Case in point was Monday night / Tuesday morning at the O’Hare Hilton, where after about a half hour of standing in line at 12:30 AM, the front desk staff tried to convince me that I didn’t have a room (this was before they told me they were “cleaning it“). Anticipating just such a difficulty, I’d tried to get on the network during my overlong stay in line, with no success. Fortunately, I’d sent my travel information via email previously, so I was able to fish the confirmation out of my Sent Mail folder in Evolution to wave in their face. In a network only scenario, I’m looking at retreating to the airport for connectivity (which the American terminal doesn’t have, shockingly) – which would probably cost me $10 despite the fact that I only needed a few minutes of access, then another 30 minute wait in line (don’t say I could have printed it out – I carry a machine so that I don’t have to lug around paper). Or take the example where Jon Udell talked recently of his ability to leave his laptop in the bag during a two week vacation. I find that even with thin-clients, I prefer my implementation of Firefox to someone else’s, because its customized uniquely to me. The network may well be my computer, but I’ll take my version of the terminal over someone else’s.
As important as it is to have that kind of cached, local access however, the increasing power of the network is undeniable. For every question I can answer with the information I capture locally, there are a million, millions probably, that I can address only with network access. Do I dream of a day where access to the network is ubiquitous and I can carry that power with me anywhere I go? Sure. But I’ve been dreaming about that day for 5 years now, and while my mobility options are better than they were, I certainly cannot say that I have internet everywhere. Nor will EV-DO or UMTS solve that for me, because as hard as it might be for some to believe, there are still places where you can’t get a cell signal. My loft, for one. Damn concrete.
But who knows, maybe Google will blanket the whole country with wifi and the point will become moot. Until then, however, I do believe strongly that the desktop does matter. But the real question, the one I haven’t seen too many people address, is whether or not it matters which desktop?
That to me is the near to mid term importance of the rise of thin-clients; they reduce the fixed dependency to a particular brand of operating system. For the record, I mostly agree with David Berlind when he says:
If you’re waiting for something to move the needle on the desktop the way SaaSers have done it for on-premises backoffice solutions (and the way it should be moved), Desktop Linux isn’t that something. Neither is the Mac. But Dana Gardner is absolutely right that the network computer is.
I think his view is a little too North America centric, as I believe desktop Linux has the potential to be huge in important geographies like China and India, if for no other reasons than cost and politics. Further, the two pieces discussed – Linux on the desktop and Software as a Service (SaaS) – are not unrelated. The rise of SaaS, in other words, is a tremendous enabler for desktop Linux because when you’re running your applications on the network, just what the underlying operating system is matters a whole lot less.
And don’t think it’s just the Google’s and Salesforce.com’s of the world making this play; there are a number of commercial entities with similar ambitions but differentiated technical strategies. IBM’s Workplace, for example, is all about making the underlying operating system more or less an afterthought; the application set is decoupled from the OS via what is essentially middleware for the desktop. Sun with its Tarantella acquisition harbors similar ambitions of delivering even rich client Windows applications to the desktop, be that Windows, Linux or a thin client Sun Ray. Are these intended to kill Windows? Not really, they’re happy to run on it. But they don’t require Windows, which is an important shift. It’s not another attempt to out-Windows Windows, but rather the old embrace, extend, extinguish play.
So to me, the really important question is not whether or not the desktop matters, it does and will for the forseeable future. No, the real question is does it matter who made the desktop you run? And let’s be honest, when we’re discussing this, does anyone really debate which desktop we’re talking about? While I’ve been very impressed with the revamp of Office [1] and the work on Vista, I find it difficult to believe that a couple of years from now that that product combination will be as important as it is today. Before people get carried away and read the death of Windows or Office into that statement, let me clarify: it’s my belief that both Office and Windows are going to be around and viable for at least as long as I’m likely to be in this business. But will they be the only game in town as they are today? As John and I discussed the other day, I can’t see how they could be.
It’s not clear to me which of the SaaS/Workplace/remote hosted applications will be most effective at decreasing the dependence on Microsoft’s desktop, as I think they will appear to different audiences, but I have to think that some of them will. Maybe not huge share, but some. Similar concerns may in part explain the recent reorganization that folded MSN into the main product unit, but I have to question whether or not Microsoft can afford to risk strong plays into the SaaS arena. Just how nimble can they be here? Their decision to not support the ODF alongside the MSXML format suggests to me “not very.” And really, who can blame them? Any even potential jeopardization of the dual licenses to print money that Office/Windows franchises have been would have Wall St absolutely howling. As Jon Udell (I think it was him – correct me if I’m wrong, Jon) said at the close of the 9/9/05 Gillmor Gang, Microsoft is hamstrung to some degree by the commitment to its major franchises. Its strength, in that respect, is also its weakness.
So while I’m certainly not aligned with some of the doom and gloom scenarios predicted for Windows – Vista looks terrific, I can’t wait to try it – I do believe that its King of the Hill status is due to be taken down a notch or two. They’ll probably still be king, but much like John after the Magna Carta. The crazy thing is that I think it will be good for Windows; much as Firefox has forced IE to respond, Windows could use someone pushing more seriously than a closed hardware system like OS X can. Whether that’s Google’s SaaS, IBM’s Workplace, desktop Linux or something else isn’t important to me, as long as the winner is choice and cross-platform solutions.
[1] As impressive as I find the technology, however, I find it even more impressive that Microsoft is willing to take such a huge risk with the UI at a time when Office is viewed as somewhat vulnerable. While much has been made of the retraining costs of migrating from Microsoft Office to OpenOffice.org, I think the case can be made that retraining on the new UI – impressive as it seems to me – could be similar in scope.
