What will it take to make Linux popular?
In a recent interview Linus Torvalds, the mastermind behind the Linux kernel, said that the operating system is not as popular as Windows on consumer PCs because it doesn’t come preinstalled. Manufacturers sell the computers they make with an operating system on board, which most of the time is Windows. Why can’t it be Linux instead?
According to Net Applications, in 2011, sales estimates have Linux at roughly 1.5 percent usage share on desktop and laptop computers. Windows on the other hand was evaluated at 92 percent in the same estimate. The discrepancy in sales points out few of the issues that Linux has to overcome in order to reach a broader market adoption, but it can also provide a solution.
Here lies the problem behind Torvalds' thinking. He believes that there is a direct relation between Android coming preinstalled and its success and that this can be mirrored by Linux by coming preinstalled on new PCs.
The problem is that Android is a mobile operating system and it is designed from the start to operate like one. The popularity of Android can’t be attributed to it coming preinstalled on mobile phones and tablets, as this is a basic assumption that it does and we expect it to work this way. Nobody would want to install the OS on the device right after buying it. It’s just not natural. The popularity of Android comes from the whole user experience, the diversity it offers in the market and the performance of the devices running it, among other reasons, but none of them is that it comes preinstalled.
On the other hand, Linux is not designed to work the way Android does on smartphones. Android is indeed derived from Linux, but the interface is not like any interface made for Linux. To make "popularity happen", a redesigned interface should be offered that would provide a similar experience to Android.
It may not be ideal for power users, but most buyers are not experienced setting up a Linux distribution; for them, ease of use is just critical. Any bump in the road can be very challenging and even considered a dead end. Linux developers must pave the road by rethinking what the market needs to open itself to alternatives to Windows.
But There is a Solution
There is some logic behind Torvalds' thinking. Linux needs support from computer manufacturers like Dell, HP, Acer, ASUS, etc. The big players need to offer the option to have a distribution (flavor of Linux if you will) preinstalled that is user friendly and at the same time can satisfy all the user’s needs (like an available office suite for instance). It has to have all the basics covered and work from there.
A very good place to start is to provide a consistent user experience that doesn’t involve the user needing to know anything related to how the operating system works. Why? Because the majority of people buying a new PC expect it to work from the beginning; they don't want to deal with installing drivers or the complexity of setting up every detail on a freshly installed operating system. The Linux community needs to treat every person as a guest into their ecosystem and not demand active membership. The guest should have everything at his or her disposal, but shouldn’t need to work too much to get it, other than taking a few simple steps.
It Can Be Done
"How could that be possible?" you ask. The idea behind this is not too complicated and it involves a simple recipe: find a rounded Linux distribution as a starting point with good community support (like Ubuntu for instance) then make sure it works perfectly on the computers you want to sell it on and then help people set it up before they go home with it. Imagine not having to do anything when you get home and just use what you just bought -- no tinkering required.
The world doesn’t need to get introduced to the hardcore part of Linux. If you want that you already have it and that’s why I think a fresh set of ideas should be implemented to pave the way for Linux to become more popular in the consumer market. This is what the people that want to make this happen should fight for.
Trying to make something that isn’t completely user friendly work is a more difficult process in the long term than just starting fresh. Trying to create a completely redesigned OS from the start is not necessary. The parts are ready, they just need to get packaged right.
What do you think should be done? Could this idea work or is it just wishful thinking?
Linus Trovald's interview at Aalto University in Finland is embedded above. Watch out for the kind words towards nVidia at the end. Perhaps you heard about them already?