Sparks Fly As Linux Kernel Guy Quits In A Huff

There's another item to add to my list of 7 Reasons Why Linux Won't Succeed On The Desktop. An Australian doctor who spent his spare time working on the kernel says desktop performance is suffering because Linus Torvalds and company are only concerned about enterprise users. He got so ticked off he quit Linux in a huff, and made some pointed comments about Linus.The doctor/developer is named Con Kolivas. His tale is pretty important, especially in the context of all the push-back to my criticisms of desktop Linux. Essentially, Kolivas agrees, but for completely different reasons. His take: "The desktop PC, for which Linux started out as being developed for, [has] fallen by the wayside. . .The developers were all developing for something that wasn't the desktop. They had all been employed by big name manufacturers who couldn't care less about the desktop (and still don't) but want their last 1% on their database benchmark or throughput benchmark or whatever."

To remedy the problem, Kolivas wrote something called the -ck patchset, which is a bunch of code that changes the kernel's CPU scheduler to improve desktop performance.

Long story short, Kolivas's fight to get his patches folded into the kernel was rejected. Adding insult to injury, another developer wrote a scheduler, which did get added to the kernel. At his wit's end, Kolivas decided to stop working on Linux, charging that the development process is completely disconnected from the needs of normal users ("You know, the ones who constitute 99.9% of the Linux user base," he says.)

But wait, there's more! Kolivas raises the very interesting point, which rings true, that the people who maintain the Linux kernel are so intimidating to the average Linux user that Linus and his inner circle never even see many bug report from those average folks. Here's how he explains this:

Which leads to his parting shot at Torvalds:

Kolivas's feelings aside, the salient question is whether Con is correct that desktop Linux is suffering because of an over-focus on enterprise users and server Linux. A sober look at the situation, from the perspective of the Linux brain trust indicates that that's probably a sensible approach.

Linux is in widespread use on servers, and there's actually a business surrounding this. Desktops, not so much.

I can't tell you whether Kolivas's contributions would have actually improved the desktop performance of the kernel, though on the face of it his ideas make sense. Whether he's the best guy to implement them is another story. Clearly, Kolivas is a polymath, and a highly intelligent one at that. (Now that he's quit Linux, his new project is to learn Japanese.)

The whole tale tells us something else, which runs counter to the image of Linux. Most of us think of Linus, the Linux folks, and indeed the whole open-source community as more fact-based than, say, the grubby commercial software sphere in which the Microsofties play.

Obviously, that ain't true. The Linux folks are just as prone to internecine warfare and backbiting as the rest of us..

Here's the link to the story, Why I quit: kernel developer Con Kolivas.