Choosing A Linux Distribution, Part I
The power to choose is a good thing, so long as you are choosing right. That has become difficult for most people in the Linux world, where there is a bewildering range of possibilities, and the advice you get, instead of narrowing things down, tends to simply re-emphasize the great quantity of your options. It’s no wonder that people who might be interested in trying an alternative operating system never start. In this two-part series, I will present a list of some of the factors that I considered in choosing a distribution, and show the reasoning process in action.
I. Is there a distribution recommended for use, or bundled, with the hardware?
For a Raspberry Pi, the most sensible choice for most is going to be Raspberry Pi OS, for the simple reason that you can’t expect better support for a Pi than from the company who makes them. If you spend a lot of money buying a System76 laptop, it makes little sense to then use some distribution besides System76’s own Pop!_OS. If you buy a Dell computer with Ubuntu preinstalled, you need a clear and compelling cause to switch away from it to something else. (Wanting to test other distributions is not one; temporary experiments with alternative forms of Linux are best done in a virtual machine.)
II. Is the distribution is likely to endure?
In line with ‘Lindy’s law’, distributions that have been around have probably gotten their act together, and can be expected to stick around. Longevity and a proven track record cannot be underestimated in projecting future reliability. Besides this, there is also safety in numbers; a distribution’s odds of survival are enhanced if it has a significant number of people, or significant organizations, behind it, in development or otherwise. A distribution with one or two developers can be taken out of action if something happens to them, but with many contributors to fall back on, the project is likely to continue into the foreseeable future. (Longevity and community size also typically have the plus of making it easier to find support as a user, a substantial benefit that could probably be its own factor here, but isn’t.)
III. What is software support like?
Established distributions tend to have larger software repositories, and developers tend to prefer those when they’re making packages. That isn’t to say that you can get software on some distros and not on others; nearly anything is possible if you have the skill and the time to work it out. However, every hoop you have to jump through is time you could spend doing something more productive. Generally speaking, you’ll have an easier time setting up software on more mainstream distributions.
IV. How easy is it to set up, use, and/or maintain it?
Some distros are easier than others. The three major problem areas with Linux setups tend to be with initial installation, driver support, and/or updates. The fact that a distro might be perfect when it is properly configured doesn’t, by itself, override the reality that it might be a tremendous hassle to get there in the first place, or to keep it from becoming improperly configured after a bad update.
V. What type of applications do you expect the distro to come with after a fresh install?
In my case, I feel that less is more; I like knowing the majority of what is on my system, and I prefer to install programs on my own instead of sifting through and uninstalling everything I never wanted.
Even after all of these technical considerations, and perhaps others, are taken into account, there still tend to be a lot of options, too many to conveniently pick from. A would-be Linux user must choose something, however. I suspect that at this point, many fall into analysis paralysis, or go into ‘distro-hopping’ mode, trying out all possibilities to see if there is a ‘single true distro’ that clicks with them. I feel the secret to overcoming this is to think of it as not just a logical decision, but an emotional one.
When there are no critical technical differences between a product and its competition, then the most relevant factor to its users is...
VI. Do you like it (whether the product, or the brand, or the people who put it out) more?
If software X ‘clicks’ and you’re at peace with it, and its competition feels ‘icky’, then you don’t need to find some technical reason to justify the choice you’re already inclined to make. It is sufficient, and practical, to simply accept that you feel better with X and move on with your life.
There are many intangibles that go into Linux distros, and which you believe are important are quite personal. I look at a wide range of things: the projects’ mascots, the community’s general demeanor and attitude (this factor may be a little more tangible than the rest, if you need tech support), stated manifestos/philosophy of the project, known facts about the most prominent developers, and the names chosen for things, whether for distro releases as a whole, or for components. (There are a lot of distros, desktop environments, or programs that I have ruled out for this last reason alone.)
After all of these considerations, you should be down to a relatively few options, enough to make, if not a permanent decision, then at least a long-lasting one. Stay tuned for the follow-up article in which I will explain how I evaluated these factors, what distribution I ended up with, and why a few other candidates were ruled out.