The new future of Linux?
Red Herring online had an interview with Eric S. Raymond this week following his remarks about the near future of Linux and open source. In short his arguments boiled down to this. Desktop computing is nearing another crucial threshold. ESR found that major shifts in operating systems during technological transitions. Like from 8 bits to 16 bits and for 16 bits to 32 bit computing. Currently, we are looking towards 64 bit computing and he expects the window of opportunity to close for years after that. The key challenge is to increase the Linux user base before this happens.
Building on that argument he looks at current non-geek users. The people who just want things to work out of the box. Who want to plug in their iPod, listen to MP3s and watch DVDs, without any additional hassle (and not run the risk of breaking license agreements in the process). ESR recommends Linux distributions not to focus on open source purity but on user experience. And add proprietary drivers and codecs to get there. He sees this as a temporary measure to push Linux to the desktop and not loose the current window of opportunity. ln the Red Herring article he specifically mentions Linspire as the distro to watch and follow.
Now, you can’t dismiss his remarks off hand. One of the first things I did after installing Ubuntu was to use EasyUbuntu to solve the issues with MP3, codecs, flash and java. So yes, usability is important.
I don’t agree with the technological argument nor the applauding of Linspire. Linspire may have integrated some of the proprietary elements and finally made CnR free, it doesn’t have the needed combination of innovational drive and usability. The Linspire distro did follow the developments in the kernel and the applications, but the desktop is still uninspiring and cluttered. And where is the 64 bit version that is needed to make use of the technological shift.
I don’t buy the window of opportunity argument. It is not an enduser argument. When computing went from 16 to 32 bits no enduser even contemplated moving to any other OS than Windows. If anything, the current userbase of Windows is even more massive. Linux did grow in other markets and it did not need a hardware shift to achieve that. On the contrary, Linux was loved because it allowed mission critical business use on generic hardware, thus lowering the costs. Linux made some inroads on the desktop, but not massively so. Why not ?
Because it still suffers from ease of use all across the board. Installing it is no sweat. Updating, upgrading and installing software is almost painless, provided you stick with the distro’s repositories. Almost. Remember the problems with Suse’s update, or the Ubuntu foul up of Xserver. Both problems could be solved, but the solutions required more than a general user would know. That is why Linux is still not massively visible on the desktop.
I remember a problem with Suse 9.2 or 9.3. It would break your ability to boot Windows XP in a multiboot setup (not uncommon for migrants). Solving it required a patch. About six months later I got the Novell Linux dvdset with the same distro. And it again broke the multiboot. The patch was not fed back in the distro, so again new migrants were confronted with a well-known problem. Hence, another year lost.
Using Windows will have it’s share of problems, but the platform does have it’s set of problem analyzing and solving tools. From the Windows restore point to tune up software. The current Linux distributions do not have those tools. When a new end-user runs into a snag he has no clue what is wrong or how to solve it.
Raymond did a good job in opening the eyes for a shift in focus. I believe he chose the wrong distribution for it and used the wrong technological argument. Companies like Canonical, Red Hat and Novell should be able to furnish for the license costs, but so far all three stay away from it. But with Ubuntu gaining momentum and Suse Linux Enterprise Desktop getting wide acclaim they might want to listen to ESR and their userbase.