Enterprise Unix Roundup -- Longhorn's Loss, Linux's Gain
|Main||In Other News||Security Roundup||Tips of the Trade|
We'll admit to a little mystification over the apparent ripple that passes through the pundit set every time Microsoft bestirs itself long enough to announce something is or isn't coming in on time, and is or isn't going to have this or that gewgaw or widget.
On one shoulder, our sour-faced cynic chirps away, "It's a big company ... like those dinosaurs with brains near their tails ... It's supposed to blunder around and miss dates and not do stuff it thought it could when the peanut brain thought the stuff up and tried to get the tail-brain to go along."
On our other shoulder, the pious long-view sage tells us, "Enterprises must be able to plan on deployments else chaos will ensue! Corporate America is poised to take up Longhorn ... It just needs the word!"
With its recent decision to drop WinFS from its next-gen OS, Microsoft caused one of those ripples, provoking an equally excited rumble from the Linux world, which (naturally) perceives the feature delays as a window of opportunity for some progress in an area that's proved difficult to colonize: The general-purpose corporate desktop.
By our reckoning, every year since 1999 has been the year for Linux on the desktop. Every year since 1999 has also been the year for explaining why Linux isn't a desktop powerhouse.
Anyone who's been watching Linux since the late '90s is aware of the massive energy that's gone into making it presentable for desktop users, both technically, in the form of hundreds of assorted software projects large and small, and rhetorically, in the form of countless exhortations about this being the year for Linux on the desktop. By our reckoning, every year since 1999 has been the year for Linux on the desktop. Every year since 1999 has also been the year for explaining why Linux isn't a desktop powerhouse.
The rationale for a while was "ease of installation," which provoked at least one Web page explaining how to manage a Debian installation with the end of a pencil. Other rationales have been tossed around, too: Bad interface, bad hardware driver support, a paucity of familiar applications, no games, and so on.
We periodically re-evaluate the state of the Linux desktop when the "this is the year for Linux on the desktop" noise gets loud enough, and each time we come away with the same basic impression: All the pieces seem to be there, especially for enterprise use. In our mind, the real headache for an enterprise looking to switch to Linux is the simple pain of testing, rolling it out, and transitioning the users. So much so that the endless stream of commentary about "what Linux really needs to rule the desktop" looks a little foolish because the answer seems to be "Microsoft's utter annihilation."
No planner ever took Microsoft's initial release dates and feature lists for Longhorn seriously, and no planner was horribly surprised when those dates and lists slipped.
Hence, the current excitement over Longhorn arriving perhaps not a day late, but definitely a dollar (and a cool metadata searching filesystem) short. The wisdom behind this excitement is that the corporate world is ready to roll forward with something, anything, no matter what, and if Longhorn can't satisfy IT decision makers' cravings for progress, then Linux will surely do.
Except it doesn't have to. Although IT spending appears to be trending up again, and the stage is being set for a veritable explosion of spending some time in the next few years, it doesn't follow that enterprise operations need to spend that money on a new desktop operating system. No planner ever took Microsoft's initial release dates and feature lists for Longhorn seriously, and no planner was horribly surprised when those dates and lists slipped. Further, most operations are well-accustomed to making do with what they have after several years of slashed budgets.
So what about that window of opportunity? We don't think it's going to be found in the length or brevity of Longhorn's feature list. It's going to be found in the cost/feature mix that Novell, Red Hat, and Sun bring forth with their desktop Linux offerings.
Sun's Java Desktop System is currently the most promising because it already plays aggressively at several levels, both consumer and corporate, in a way the other two's offerings haven't managed yet. It also helps that Sun is desperate enough to make its licensing fairly generous. Red Hat, on the other hand, is confident enough in its dominance of the enterprise Linux server space to give its most recent desktop offering a fairly high price tag. Novell is still something of an unknown quantity. The value of its SUSE offering is still to be determined as we wait around for a look at its actual desktop software and how well it manages to tie that in to GroupWise and NetWare at large.
Presumably, one of the three will look past the issue of whether Microsoft is perceived to be "stumbling" with Longhorn and concentrate on delivering value and features. And that's where the real window of opportunity is.