Will demand for quick-booting PCs give Linux a leg up on Windows? That's how some folks see it, but I'm not one of them.A feature story in last Sunday's New York Times set the stage for this discussion by focusing on a sore spot for many PC users: glacial PC boot-up times:
"It is the black hole of the digital age ï¿¼ the three minutes it can take for your computer to boot up, when there is nothing to do but wait, and wait, and wait some more before you can log on and begin multitasking at hyper-speed." As the Times story points out, many top-tier PC makers are rolling out systems that deliver quick boot access -- defined as 30 seconds or less -- to email, a Web browser, and other basic apps. Many of these systems use some version of Linux for quick-boot functionality, either in place of Windows or as a stand-in while Windows starts up.
Jim Zemlin, executive director of the Linux Foundation, sees this trend as the first temblor of a seismic shift in favor of desktop Linux:
"We may see a world at the end of next year where Linux ships on almost every notebook computer regardless of whether it is loaded with Windows. This in addition to the huge potential of the netbook, mobile internet device and mobile Linux market can mean huge and immediate inroads for a Linux desktop, albeit not in the form most people had predicted many years ago when the first ï¿¼year of the Linux desktopï¿¼ was declared."
I won't take issue with the idea that Linux is in many ways for better suited than Windows for use on netbooks and on many types of non-PC Internet-enabled devices (including, of course, smartphones). I think it's more of a stretch to assume that these OEM quick-boot schemes will really qualify as meaningful market-share gains for desktop Linux.
The real problem, however, is the idea that this approach will make users either more aware of the benefits of desktop Linux or more favorably inclined towards Linux. In fact, rather than saving the day, these schemes may taint Linux unfairly with a whiff of guilt by association.
First, let's consider why many PC makers are implementing Linux-based quick-boot systems: instant gratification. I see this as a case where OEMs are likely to raise user expectations and then fail miserably to deliver the goods.
How, for example, will Linux and Windows interoperate on a host system? Something as simple as cut-and-paste functionality between the Web browser on a quick-boot Linux OS and a user's Windows-based productivity applications could quickly make a mockery of the whole process.
I have more questions about how, exactly, this process might apply to desktop systems. A quick example: The biggest source of bootup delays on one of my systems comes courtesy of the RAID BIOS, which takes about 20 seconds to initialize. Still other slowdowns, such as a BIOS configured to run a full RAM test at every startup, are easy to avoid -- if users are willing and able to change the BIOS settings that control them.
Consider also how many Windows startup problems happen after a user logs into his or her desktop account. I routinely yank startup processes whose developers arrogantly assume that I want their apps to run every moment my system is running. In many such cases, less experienced users blame Windows for running slowly -- in some cases so slowly that the UI fails to respond for a minute or more after logging in -- when third-party applications are the real culprits.
Finally, I have questions about how these quick-boot schemes will help users who find that restoring Windows from stand-by or hibernation is a constant crap shoot, often with lower odds of "winning" than they would get at a Vegas casino table.
There is a lot to like about a modern, mainstream desktop Linux distro, especially on today's netbook systems -- and especially with killer distros like the latest version of Ubuntu hitting the street. But using Linux as a band-aid fix for more serious, deeply-rooted problems not of its making sounds more like a recipe for trouble than a formula for success.