In recent weeks the company has announced that it would be replacing GNOME's default UI in its next release, and has embraced a replacement for X on the desktop called Wayland. Canonical have been putting much of its development muscle behind Ubuntu One in the last few releases, and on a lot of polish for the desktop look and feel, and they've been replacing many desktop applications with simpler software in an attempt to ensure that the Ubuntu desktop is suitable for mainstream users. But all of this isn't worth diddly if they can't get Ubuntu in front of more mainstream users — and Canonical is in for a long string of disappointments if they're hoping Dell, HP, or any other major OEM will back Ubuntu in a real way.
Even though Microsoft has been showing some weakness of late, it doesn't mean that OEMs are likely to back a contender that doesn't pony up the co-marketing money that Microsoft and Intel shovel at hardware partners to retain their preferred status.
Pulling OEMs in the direction that Canonical wants to go is going to be slow, expensive, and probably fruitless. It certainly has been so far. Having watched the major vendors for the better part of a decade, I'm convinced that they're totally unsuited to do anything but ship Windows desktops. The marketing departments don't understand — nor want to — Linux. The management is unwilling to invest serious cash in promoting Linux. If you ask the right person at one of the vendors, they'll say that they've given Linux a chance and point to a few token systems buried on the Web site -- and then complain that "people just don't want Linux."
Bottom line? Canonical has already come to the conclusion that prodding the open source community in the direction it wants to go is too slow. They need to come to the same conclusion about the OEMs, and quickly.
Of course System 76 isn't the only game in town, but it would be my odds-on favorite for Canonical to purchase. Unlike most Linux-selling OEMs, System 76 made the choice early on not to diversify and offer several choices of distribution. They've got a good line of 100% Ubuntu systems and have been slowly growing over the last several years. (Note that I also think very highly of ZaReason, if you're looking for an Ubuntu system.) Canonical could start from scratch, but that would be a waste of time and money. Better to pick the company with the most affinity for Ubuntu and similar mindset. I've spoken to System76's president Carl Richell a number of times over the years, and he's got the same Jobsian perfectionism and vision that Mark Shuttleworth has and that a Linux OEM needs.
With its own hardware vendor, Canonical could focus on making Ubuntu work perfectly on a small set of systems. The company wouldn't need to stop supporting generic hardware, but it could (as it does with other things) leave much of that to the Ubuntu community and simply offer a line of supported machines. And I don't just mean desktops, laptops, and netbooks. Canonical could also offer top-to-bottom solutions for small and medium-sized businesses based around their server solutions and desktop machines. The company has been investing quite a bit in professional services over the past few years — it'd probably be a very good thing if they could simply offer a whole stack of hardware and software.
Convincing people to install Linux one system at a time isn't getting it done. ("It" being fixing Ubuntu's Bug Number One.) Trying to wrangle deals with the major OEMs isn't getting it done either. It's time for Canonical to take the plunge and start selling its own line of systems, and the fastest way to do that is start by buying a hardware vendor that already specializes in Ubuntu systems. Though I don't control the pursestrings at Canonical, I hope that the Powers That Be are planning or start to plan controlling the hardware as well as the rest of the Ubuntu experience, rather than leaving it up to companies that don't understand or care about Linux.