Sheena S. Iyengar was studying Social Psychology at Stanford in the 1990s. She often visited the local market in Menlo Park, packed with all sorts of varieties of produce. She 250 types of cheese, 75 types of olive oil, 150 types of vinegar, etc. Iyengar he realized one day that he didn’t buy anythingand that led to an idea that he wanted to validate, so he did an experiment with one of the merchants.
In one of the stalls he alternated an offer of 6 jars of jam with another offer with 24 of those jars. After several shifts, he came to a striking conclusion: the 24-jar table attracted more people, but the one with 6 jars prompted more people to buy. And that, dear readers, is a bit of what happens to Linux.
Fragmentation killed Linux
The story, which I also mentioned in an old post on my other blogis not mine, but of The Chronicle Review. Iyengar would end up writing a book called ‘The Art of Choosing’, and in it he discussed the paradox that although choices are good, having too many options can be counterproductive.
It is of course what seems to happen with Linux, I argued then. This operating system is fantastic and is present everywhere, but in most cases his presence is almost invisible for the user. It is the heart of our Android mobiles, it is the basis of the operation of the internet and the cloud and absolutely dominates the segment of supercomputers.
And then there is the eternal joke or meme of “this is really going to be the year of Linux on the desktop”. It never is: Windows (87.56% according to the latest data from NetMarketshare) and macOS (9.54%) clearly rule our PCs and laptops, and Linux is still fiddling with a market share of 1%.
It’s never gotten past that and probably never will. That’s what I was talking about An old periodical linuxero, Steven J Vaughan-Nichols, who has been writing about Linux and Open Source since the war was not between GNOME and KDE but between Bash and zsh. I’ve been talking about this operating system for almost as long —one of the first articles I earned money for was a topic about how to install linux on the amiga—and I know very well what Steven is talking about.
there is a paragraph especially successful that explains a little the situation that Linux has been going through for years:
We have many excellent Linux desktop distros, which means that none of them can gain enough market share to make a dent in the overall market.
That’s how it is. The famous Distrowatch website, which has been used for years to follow the market for the various Linux distributions, shows that today there are 271 active. Imagine having 271 versions of Android or Windows, each one essentially the same but more or less different in their way of doing things. In Wikimedia you have one of the latest versions of the Linux distributions timelinean amazing graph that shows how those Linux “jam jars” have been created.
There are also 21 desktop environments and half a dozen package managers, plus new application packaging systems for easier installation. Linux users are usually aware of several of them, but not all of them. Imagine what it is to face that diversity for a new user. It’s crazy.
It’s the jam jar problem. There are too many, so often one doesn’t know which one to choose. The options are great and Linux is proof of that, but those options also doom it when it comes to trying to become popular on the desktop. There are so many that one does not know which one to choose.
In fact, the normal thing is that you try jars of jam and enjoy them for a certain time and then try others. On Linux that is called “distro-hopping“. One day I use Ubuntu, another Fedora and another Arch. And then more and more. It’s great for OS lovers, but not so great for those who want something more or less “universal”.
There is nothing really “universal” in Linux, because if this operating system is defined by something, it is by its wealth of variants. What do you not like a distribution? No problem. Wait: do you like the distro, but not its desktop environment? Install another one and pull miles. And so with almost everything.
To succeed on the desktop, perhaps the only option would be for the big distros to join forces to create something like the unified Linux distribution. A single reference distribution, a single desktop environment, a single package manager, and a single app store. And even then, I think, success would not be far from assured.
But watch out: maybe it’s better this way. It probably is, after all. It’s okay that this isn’t the year of Linux on the desktop. It is almost everywhere else.