Members of the open-source community are expressing concern over rapid feature changes in the Linux 2.6 kernel, which they say are too focused on the desktop and could make the kernel too large. Sam Greenblatt, a senior vice president at Computer Associates International Inc., in Islandia, N.Y., said that while the kernel is evolving for the desktop, server and embedded markets, more and more technology is being included, and the kernel is "getting fatter. We are not interested in the game drivers and music drivers that are being added to the kernel. We are interested in a more stable kernel." Read the rest of the article at eWeek.
Interesting article... basically, it seems to me like Greenblatt's saying, "these aren't the features we [Comp. Assoc.] were hoping for, instead they've included a bunch of other stuff" in such a way as to sound like the kernel is out of control, getting bloated and unusable. This is where it starts looking like FUD... Anybody who's got much experience with Linux will tell you that the first thing you do is compile your own kernel. This is significant because the Linux kernel is modular; you pick and choose the features you want. You don't have an EMU10k1 chipset you say? Don't include that module. You do have a VIA chipset? Do include that one. It's as simple as going through a quick list with self-explanitory entries. Most kernel options say implicitly "If you don't know what this is for, you probably don't need it!" encouraging the system administrator to take the initiative to run a lean kernel. Why include features you'll never use? My kernel, for instance, is a 2.6.8 kernel (with a few vendor-specific patches) which I compiled myself. I don't have a single thing compiled into my kernel which I don't need, so it's fast, lean, stable & secure. More features are good, if they're optional. With the Linux kernel, they are optional. -AT
