In an article entitled "Desktop Linux? Stick A Fork In It!", InfoWorld blogger Randall Kennedy tries to expose community development and control as a myth of open source software, or at least as being more inconsistent than some would think.
In many projects, the community does rule, and the community development model rocks. But as you get into older, more complex, and more "important" projects like XFree and the Linux kernel, you run into a core of hardcore keepers of the faith who essentially control the project. When you have a great idea for a feature or improvement, and do the work to make it work, and build up a large mass of users who are finding your work a great boon, you still have to get your work approved by this techno Council of Nicea before it becomes part of the official source tree for the project.
Kennedy laments how some core scheduling algorithm patches by Con Kolivas were never accepted into the Linux kernel source tree. This finally caused Con to throw up his hands and quit. Kennedy further claims these algorithms improved Linux performance for desktop uses and their failure to get accepted into the official kernel showed bias against desktop Linux by Linus and his lieutenants.
Kennedy says... "the fact remains that, at the kernel level at least, Linux is still controlled by a small group of elitist 'prigs.' And if a particular feature or function isn't a priority to them, it isn't a priority for Linux as a whole."
What he fails to note is that the kernel is the the most rudimentary and important part of the OS. Consider it sort of like the autonomic nervous system (ANS), the part that keeps your heart beating, your lungs breathing, your eyes blinking, and your glands... ummm... glandulating, all without you having to consciously think about it. Fight or flight reflex? ANS. Those unpredictable and uncontrollable boners you got in Junior High? ANS. If your ANS isn't working just right, it can cause anything from minor irritation to major death.
On the other hand, he brings up an interesting point for discussion... forking Linux into desktop and server flavors, so the priorities of different needs could be better met in the modified kernels rather than a "one kernel fits all" model. I'm not Mr. Linux, so my question is, would creating a desktop Linux fork be significantly different from what's been done with embedded Linux? Is embedded Linux even a fork?
But back to Kennedy's "prigs" comment. It could be claimed that Linus and his lieutenants have gotten a little staid and risk-averse as Linux has gone from computing foie-gras (a luxury savored only by those who could afford and appreciate it) to computing potato (a staple that eventually finds its way into everyone's diets in one way or another). But that can be because Linux has become such a large part of the computing food chain. A little tainting will no longer give a few uber geeks diarrhea, but can cause millions of servers to barf and fall over.
Where Linus and his lieutenants should draw the line is debatable, and much like a religious issue, the views can be wide ranging and their proponents frothing at the mouth, ready at any moment to fall to the floor, begin convulsing, and start speaking in BASIC ("poke 33,40... goto 10"). But calling them "prigs"? Okay, it's entertaining. But if we start with "prigs", can "doodyheads" be far behind? The debate is inevitable. The name-calling doesn't have to be.
Entries (RSS)