Even Debian doesn't let you switch init systems easily - sysvinit is marked 'essential', and you can't uninstall it. All this talk of Wayland, I don't see people up in arms that Wayland will be bad because everyone is adopting it and our choice is being limited. There are other parts of the stack that are hard and also chock-full-o-effort, like the Display Manager. because it's a hard problem, and a lot of effort. Arch tossed replaced sysvinit by default, and decided that for the sanity of packagers, they wouldn't go to the trouble of maintaining multiple init systems. I'm not going to say the description is wrong, because it's true. Contrast with the kernel, which is developed commercially, but essentially steered by a community effort - with corporate sponsorship, so not a perfect example, but whatever. RedHat not only contributes, but steers GNOME. This has a natural consequence on the direction of projects when the steering is also part of the funding. The volunteer community goes home and works a little bit. They're there, 8 hours a day, contributing. Funded development progresses faster than volunteer development. Related to the first point, follow the money. See, for example, the recent Linaro Keynote speech that GKH did. We need to help educate companies better that yes, they can make money without vendor lock-in, even when they operate within the community instead of the classic 'toss-over-the-wall' style. Companies want to get in on it, and this is, for the most part, the corporate style. Point the first: I don't think we need to call it a corporate conspiracy - this is a sad consequence of Linux' success. I liked this article and found myself nodding my head along with it.

Ratpoison arch