The text of this article originally appeared in the December 9, 2004, issue of SuitWatch, a bi-monthly newsletter written by Doc Searls, Senior Editor of Linux Journal.

The news cleared up yesterday morning. After a week of rumor, which followed months if not years of speculation, IBM announced its decision to sell its PC Division to Lenovo, a Chinese manufacturer. The index page copy on IBM's site read: "Breaking news: Lenovo to acquire IBM PC Division, creating new leading PC business with global market reach". This copy jumped to a page with this text:

Here are some of the details gleaned from the press release:

It's clear why IBM is doing this. Desktop PCs have become low-margin commodities, and laptops, although higher-margin items, are relatively low-volume products. Manufacturing already is done offshore, and so is most of the sales action at this point. Lenovo can scale and manage the manufacturing, and with its equity stake, IBM can keep a piece of the action.

Still, I'm disappointed. I love the ThinkPad. I'm writing this on the Emperor Linux Toucan, which is the same IBM ThinkPad T40 to which we (Linux Journal) gave Editors' Choice and Product of the Year awards this past year. Using the T40 often is a PITA (look it up if you don't know), but the rewards are worth it. The screen and keyboards are killer. The trackpad and pointer (it has both!) are allied with five buttons, including the middle button that's so useful in Linux and UNIX. There isn't a more efficient and useful keyboard layout on any laptop, by anybody--in my opinion, anyway. The little nightlight that shines down from the lid is perfect for doing astronomy with KStars and for taking notes while somebody lectures in a dark room. I would like a quieter keyboard, but the action is so nice and so positive, that I have to say the clackety sounds seem more a feature than a bug.

Will it last?

Well, although I rarely expect Good Things out of an acquisition such as this, I see some hope for success with this one. Mainly because I think a Chinese center of gravity in the PC division may help pry the ThinkPad design team away from its exclusive fealty to Microsoft.

See, while everybody bemoans the commoditization of PCs, nobody talks about the real problem underneath that commoditization, which is a proprietary monopoly that imposes innovation-stifling restrictions on hardware OEMs.

This became clear to me two years ago, at Comdex 2002, which turned out to be the penultimate Comdex. As I wrote in "A Losing Bet: the Last Days of Comdex":

It was clear that Comdex had become, essentially, a Microsoft show. The biggest booth by far was Microsoft's. And Microsoft was the biggest presence in all the featured hardware OEM booths. In her keynote Carly Fiorina of HP talked up the TabletPC and avoided Linux, no doubt out of respect for Microsoft honchos sitting up front.

That year, I got the clear impression that Comdex had been hollowed out by the company and the operating system that ran everything. Room for innovation by the OEMs was minimal, both in their products and in their booth promotions.

Now the same thing has happened to the whole business.

Early this year, IBM folks were telling me and Jeff Gerhardt of The Linux Show, where I'm a weekly gang member, that a for-real Linux desktop and laptop were in the works. But over the summer, I had a talk with Dan Frye of IBM in which he made clear that IBM was not pushing ahead aggressively with those plans.

HP came out with one Linux laptop, a mid-level unit for enterprises and small businesses. (See the January 2005 Linux Journal for a review by Don Marti.) Nothing yet, though, for the early adopters who drive market movement, namely, Linux geeks. Although more Linux laptops are kicking around these days--the percentage of Mac OS X laptops clearly was down a bit at November's Apachecon, for example--no hardware company has stepped forward to drive Linux desktop or laptop sales in a serious way. They might build to order, but they're not going to make a real market push.

They'll say they're "waiting for demand", but that's bull. At PC Forum in March 2003, I asked Intel COO Paul Otellini why Intel didn't release Linux device drivers along with ones for Windows. He pleaded absence of demand. I replied:

He said, basically, that customers are "wed" to the Windows desktop. What he didn't say is that hardware suppliers are wed to Microsoft as well. And as long as the rest of the world depends on that marriage, there won't be any room for innovation on the desktop, except by equally proprietary alternatives such as those from Apple.

Since that exchange, Intel quietly has added more Centrino device drivers. And on November 24, 2004, Steven Shankland reported in CNET that "Intel has begun an effort to make it easier for sales partners in China and India to sell desktop computers running Linux, starting a more active phase in the company's help with the open-source operating system." Specifically, he adds, " 'The chipmaker warmed up to desktop PC makers when partners in the Asian countries started requesting more help with desktop Linux,' company spokesman Scott McLaughlin said."

Now, when Intel ships the components out of which companies assemble PCs--often called white box systems because they're from companies with little-known brand names--it includes a kit of software and instructions to ease Linux installation. It's a strategy Intel has used for years with Windows. The kit includes driver software, which enables use of specific hardware features; scripts to install software quickly that has been validated to work with various versions of Linux; and a program called the Application Version Compliance Tool that checks to make sure programs are compatible with those Linux versions and Intel electronics.

Note that domestic makers of non-white boxes didn't make the same request. And those are the innovative ones, right? Who more than IBM is innovative with computing, fergoshsakes? Yet they're avoiding innovating with Linux laptops? Why? Only one reason, seems to me.

So maybe that will change.

On the discouraging side, I remember when GE unloaded its radio business on Thomson in the early 1980s. At the time the company was making some killer radios, including the legendary Superadio and the first boom boxes and personal (Walkman-like) radios to feature digital tuners. Those were terrific too. But, after doing a great job of marketing a generation or two of the radios, Thomson bailed. I don't know why, but I also don't doubt it was because it wasn't their business in the first place.

Lenovo may be different. Maybe the best IBM PC division employees won't leave when they suddenly find themselves working for somebody else. Maybe they'll welcome a freedom they didn't have before. I dunno.

Whatever the case, sooner or later somebody is going to make and market a serious Linux laptop. I'll be at the front of the line, waiting.