The newest word from the Canonical camp about future features for Ubuntu Linux is booting to a desktop in, get this, ten seconds. If they can do it, great! But I suspect another reason might be to do an end run around the rather flaky state of power management in Linux.

First off, I'm well aware that there are plenty of documented reasons why fast boot is useful apart from the aforementioned. But I've tried out various Linuxes on various machines -- desktops, "luxury" and "economy" size notebooks, netbooks, you name it. One of the most consistent problems was power management. On some machines, S3 sleep and S4 hibernate worked fine; on others, they glitched; sometimes it depended on the video drivers or BIOS in question.

So it makes sense to just avoid the whole problem entirely. Boot clean in mere seconds and you also extend battery life, since you don't need to trickle-discharge a battery for a system that's been shut completely off. That's been the philosophy behind Moblin's rapid boot time, too, as far as I can tell: no reason to implement hibernate or break your neck getting S3 suspend to work on a congeries of different hardware when you can just do this.

They actually don't have far to go to bring it down to 10 seconds, but it's always that much more difficult to whittle off that much more time. My current install of Fedora 11 (64-bit) boots and gets to a desktop in about 30 seconds, sans the time required for a login. To bring it down to ten would probably involve creating and loading into memory images for certain system states at boot time -- something like what hibernation itself does, but in a more segmented and flexible way. (I also imagine they have to find a way to do all this without also impacting shutdown time.)

A while back I wrote about the first Linux boot-time shave-down project (which was for Ubuntu as well), and noted then that improved boot time was only one of several things people could glean from this project. The same rule still applies: fast booting is actually a by-product of a number of other goals. But I hope one of the other by-products of this isn't the complete avoidance of fixing extant power management issues.

InformationWeek Analytics has published an independent analysis of the current state of open source adoption. Download the report here (registration required).

Follow me and the rest of InformationWeek on Twitter.