Neat article in Forbes:
“It’s quite hard to make a big difference in an OS, since the OS itself isn’t supposed to be doing much,” [Linus Torvalds, inventor of the free Linux operating system] says. “It is supposed to be this interface between applications and the hardware, and since people have been using computers and operating systems for a long time, there’s seldom a whole lot of new things that are worth doing. Many fundamental os issues were mature technology decades ago. That’s one reason you’ll find Microsoft and Apple competing in the looks department.”
Change, though, has come to Linux and other operating systems. But it has come on little cat feet, often prompted by changes in hardware. One example is storage. Disk drives are getting huge, and while they remain relatively slow, they are often supplemented with speedy flash RAM. Small, hypertechnical changes are needed in the Linux kernel to deal with each of those new components. Changes like this: “Data structures that map the extent of blocks, rather than enumerating each block mapping individually,” explains Torvalds, though of course you probably already knew that. ^
Linus is talking about the technical details of the operating system. To most of us, however, the operating system is much bigger. Does it have drivers for our devices? The interface and GUI — how well do they work? Is there a support contract for it, or information I can get online? Is there software for it?
The basic idea of the operating system has not changed since the 1960s. As our hardware has grown more powerful, we’ve added abstraction layers to our methods of coding so that we can be more flexible and re-usable, which this hardware has made necessary.
However, what I pulled out of this article was this idea: if we’re able to finally get a nice stable platform that has all the technical features need to support what we build on it, and is so rock-solid it never crashes, then we can go to work on the interface and user experience, which even on Macs need help.