Keep It Simple Stupid Linux, Kiss for short


I’d like to inform you about my new Linux distribution called KISS.


Independent (not based on anything else).
Source based (like gentoo or crux).
Busybox coreutils.
Busybox init (busybox includes runit style services, sysvinit style services or plain shell scripts)
Musl libc.
No dbus.
No pam.
No polkit.
No systemd.
No gettext/intltool.
Packages compile without Telemetry/tracking.

Dylan Araps

 ██ ▄█▀ ██▓  ██████   ██████
 ██▄█▒ ▓██▒▒██    ▒ ▒██    ▒
▓███▄░ ▒██▒░ ▓██▄   ░ ▓██▄
▓██ █▄ ░██░  ▒   ██▒  ▒   ██▒
▒██▒ █▄░██░▒██████▒▒▒██████▒▒
▒ ▒▒ ▓▒░▓  ▒ ▒▓▒ ▒ ░▒ ▒▓▒ ▒ ░
░ ░▒ ▒░ ▒ ░░ ░▒  ░ ░░ ░▒  ░ ░
░ ░░ ░  ▒ ░░  ░  ░  ░  ░  ░
░  ░    ░        ░        ░


Update 02/22/2020

by Oneirosopher:


Hi yee all,

I’m not in charge of KISS Linux, just interested by this distro –which I have not installed by the way.

The link to KISS Linux is broken. The former site seems abandoned. Please now follow:

It seems that the same Dylan has also conceived neofetch & sowm


15 thoughts on “Keep It Simple Stupid Linux, Kiss for short

  1. This looks really promising, will s6/66-tools be supported in the future? Since it has no dbus, I assume it doesn’t include pulseaudio, jack or even sndio for apps that don’t support alsa. What’s wrong with gettext/intltool?

    Last question which boots/starts faster; s6 with 66-tools, runit, busybox or sysV init?


  2. The first part of the question is best directed to Dylan Araps himself the author of KISS.
    In my experience with pulse I have used apulse and haven’t had a need for pulseaudio anywhere. Remember that everything sound related on linux is based on alsa. Pulseaudio wouldn’t work without alsa.

    The 2nd question about booting speed is very specific in boot time, a parameter that most consider of minor importance. A few seconds more or less in booting time is not a concern when choosing an init system. Have you ever used arch live? I was surprised that it could boot so fast having systemd in it. The issue is what services are also started and how stable the system is after starting them all. How responsive and quick the system is after booting and how many resources have been used to be up and ready. It is my opinion and experience that there is nothing faster and more stable than S6. Unless you cut the power off the system is indestructible with S6. Runit is quick, and simple, but using scripts everywhere and for everything it takes time for scripts to be read and executed, and this is where binaries can increase the speed. Sysvinit works with scripts but it is also huge by comparison, therefore slow to start. Busybox is just simple and adequate for small personal systems.
    It would be interesting to standardize the same exact situation, same hw, same services, and do a comparison of what you say, but hardly anyone would choose an init and svc management based on this performance. All I can say is the 66 has made s6 booting probably as fast as possible. 66 doesn’t run while the system is active, it just sets up S6 to do its work. Only if you were to change the setup, stop/start services or modify the way they run would 66 come into play. It is otherwise pure S6 as it comes from

    I would say that if you are not interested in theoretical studying of booting and running systems and you want something quick and easy to learn to manage use runit. That is what runit does best. Small, quick, easy. If you want to have a server that you boot and you want it to run for years, s6 comes first by far.


  3. When it comes to systemd, ubuntu, manjaro, … this comes to mind:

    “Only two things are infinite, the universe and human stupidity, and I’m not sure about the former.” Albert Einstein


  4. Boot time is of no consequence for a workstation/server OS, so long as the system can be brought up in a reasonable time. So for example under or just over a minute might be reasonable, 10 minutes less so… but it varies. I for example am never in a hurry to boot an OS, others may be foaming at the mouth and climbing the walls if boot takes over 20 seconds.

    Once you consider all the variables, no two systems being the same nor starting the same daemons or having the same use cases – it becomes an apples and oranges thing.

    For “consumers” however, in this new age of instant gratification, fast boot time is often seen as the be all and end all – and for many is incorrectly interpreted as a benchmark of OS performance – that’s an old Windows era thing, where users would often rate their hardware by how fast it could bring up the system.

    As an example MS improved their boot times in Windows 8.1/10 by suspending to disk by default. For the end consumer that is “faster” and that’s all that matters – while it seems like a hack, the reality is that it’s the clever way to do it. The actual mechanics of it don’t really come into it for most.

    MS Windows is also noted for loading up a user login screen and allowing users to login while certain less critical Windows services are still loading in the background. So while you can theoretically get to a “desktop” faster, you’re working with a sluggish system, until those services finish loading.

    For something where boot time were “mission critical”, then one would imagine that the mere fact that the system has been rebooted is the actual “mission failure” in itself, rather than boot time. I would be interested to hear the use cases however.

    So called parallelised startup is similar in it’s goals, to “suspend to disk”, but obviously involves a re-implementation of part of the init / service control process. But there is the danger that you can achieve a faster boot at a cost to simplicity and of course security. But certain developers pat themselves on the back for this of course.

    Once someone removes things like rc scripts, plain text logging and move to binaries and binary logs, it becomes very transparent as to what that someone was really trying to achieve and what that project was conceived around.


  5. On a system that boots once, maybe twice a day, is file checking (root and all mounted partitions) absolutely necessary, and how many way are there in checking the file system for this purpose. I have noticed variations among different systems where 10 mounted partitions will take a few seconds to run through (and produce output of being ok, which I assume each was complete and it is not just report of the script being run) and some really slow ones, where % of a non-continuous filesystem is reported (slow=20-30sec). And I am speaking about ext* linux fs.

    Along the lines of the instant gratification centered consumer, within a consumerist society, there are all those silly reports out there comparing boot times and what they consider boot time is between pressing a switch or selecting a bootloader entry and the time they can begin entering their username in a DM, or even having an autologin and seeing the actual desktop and mouse pointer. Having a system boot up doesn’t even include having a tty1 to login, let alone starting X, a DM, verifying username-autologin, and loading all the graphical parts of a desktop. So they may have a blistering fast booting system with tons of crud that may encounter various bottle necks to end up with a funky mouse pointer.

    Systemd will claim of features, and of being able to run various tasks during boot in parallel, then the distribution may have as default 1.5′ time out for making a connection so the user seats and stirs at a timeout clock on a computer that may be meant not to be connected (like a laptop on the road with no 3-4G or wifi). So you will hear of people saying their ubuntu boots up fast but Trump forbid if you pull their eth-cable out.

    Skarnet has a history of their server performance, a log of down time and rebooting. I don’t think they remember how long it took for their server to boot because it has been many years since. I once tried killing my system (s6) or tripping it to mess up intentionally, to see what happens, it kept coming back to where I started right after boot. I did less intrusive stuff to a very simple systemd system and all hell broke loose, I had to pull the plug at the end.


  6. Scheduling of fsck really depends on the OS, the use case, the individual needs, the file system used, whether you care about the data on those partitions, etc – there is no single conclusive answer.


  7. The 2016 referendum was really a battle between two groups of elites: Those with business interests, which are entangled with EU membership and those with vast wealth who stand to gain by leaving the EU and forging trade ties elsewhere (i.e. with the US, etc). There are always vultures who will gain out of what, for the majority, will be catastrophic.

    While I’m no fan of the EU, I’m also in the position where I have to work for a living… the people at the top pushing for Brexit don’t.

    There has, as ever, been a lot of the usual skullduggery and downright lying. There were the original, ludicrous claims from the Brexiteer groups which preyed on people’s prejudices, fears and xenophobia – what we saw at campaign time was downright racism and fearmongering, since swept under the carpet and conveniently forgotten.

    The remain side, propaganda failed miserably – hence Cameron’s resignation – as it was far too patronising and was “dubbed project” fear. It probably helped the Brexiteer case.

    You should note that there were two Brexiteer campaign groups. Leave.EU, the group backed by two millionaire businessmen – found several million in funding from who knows were and is linked to UKIP and the Brexit Party. They famously breached spending restrictions and were fined.

    The other group – “Vote Leave”, was backed by several businessmen, financiers and politicians. It is effectively “running” the country at the moment. It was behind the massive lie printed on the side of a bus and has also come under scrutiny over it’s spending and was found guilty of overspending last year.

    While there were more left wing proponents of Brexit, the whole “leave” movement was decisively owned by the right wing.

    Whatever side of the fence someone may be on over Brexit – there are a few groups of elites who are desperate to get the UK out of the EU no matter what.

    We have had a long period of uncertainty and dithering since Article 50, we have yet more uncertainty to come. As I’ve said, the Machiavellian skullduggery is breathtaking. Johnson’s current strategy of shutting down parliament could be yet another such tactic. It is likely another attempt to blame shift to the EU and MPs, to remove any hope the EU may have of MPs stopping Brexit and to try and force a compromise.

    The usual Tory mantra is something along the lines of “The British people are sick and tired and want us to get on with it[blah]”. Whereas the reality is the only “British people” they care about are the landed gentry, aristocracy and super rich business people they represent.

    Personally I’m sick and tired of /them/ and want a general election, which gets them out of office and reverse the Brexit farce. And Brexit has been an embarrassing farce from the start – the idea to leave the EU, Cameron’s political gamble, but with no actual plan of how to actually do it, followed by a few more years of dithering, frantic finger pointing, blame shifting and back stabbing. And now in 2019 – after watching May’s premiership get it badly wrong, we now have “the clown” who can probably do no wrong. The years of dithering and nothingness now almost give Johnson the clear mandate for the “no deal” exit, from which he can walk away unscathed. As with Cameron, these people are happily gambling with others livelihoods and well being, because they can afford to.

    Welcome to the fabled world of 99% of the wealth being controlled by that 1%… while 99% of population share the remaining 1% of it.


  8. I’ve just tried it out. (I’m a total noob about how to install packages from outside of it’s repos)…
    I think I’m falling in love…


  9. There is the anti-capitalist left and the not-so-anticapitalist left, the left of confusion and parliamentary marketing for anything that is “moving”. I believe the position of not joining the EU is very different from the one wanting to leave it. That other left seems to have missed the boat of having a serious and complete proposal as to not joining a “common market”. The European Economic Community seems to have been just that, a common market where the largest corporations in each sector can have it all and the lesser ones would vanish. And this is what happened, the wealth was concentrated in the centers of the large multinationals’ bases, and the political power followed suit. In some cases in the EU there was a tremd of the reverse, where the concentration of political powers rolled a red carpet for their own corporations to roll all over the competition.

    The longer this carnival lasts the more of centralization and its ill effects we will see. The UK is a special case where some may have fallen victims of their own myths and rhetoric. On one side being cautious and not joining the Euro, being able to control their local economy, maybe speculating on the long term failure of the Euro, lost both in competition and in currency gamble.

    My point is that this brexit/no-brexit/deal-nodeal crisis is what the sharks need as an excuse to prey on the privileges, rights, and common/public gains of the masses. And this they will do despite of the final outcome. Thatcher’s dream will finally materialize and complete itself. And this leftleave instead of focusing on the attempt to attack the welfare state, the labor laws, and the public lands and resources, is speculating yet again on the membership status of their “empire”. It costs to keep an empire going and the cost must be transferred to the historic losers of the economic system.

    The point is not to make living in capitalism better but to overthrow it all together. I suspect the leftleavers see National Capital (and capitalism) regaining its pre WW1 triumph, in which having socialized part of the empire it would be a benefit to their voters. Being on drugs is one thing, being on expired drugs is even more alarming. After Syriza the global social-democracy left is both confused and incapable of participating in “real politics”. No other political group has been more effective in implementing brutal neo-liberalism as the “coalition of radical left – SyRizA”. Their existence does more harm than provide any hope. And this is of international significance.


  10. Regarding scripts vs binaries and runit vs s6…

    1) You have to parse and interpret the desired results regardless of whether you use a new language or use a shell scripting language

    2) Some of the performance overhead you see is more due to using bash than a POSIX shell like ash or similar. Many shells are faster by leaps and bounds. It’s why dash, even if it’s legitimately a red-headed stepchild in the space, was implemented.

    3) If you’re not using bash for init start up (Regardless of whether you’re using it as a command line shell) and you keep your scripts simple (Most attempts at runit conversion/implementation foolishly go and grab the sysvinit scripting. Yeah, it makes it faster, but parsing a 3-6 line script (which is what most run files SHOULD be…) is at the same performance level and is clear as opposed to the other.)

    As to whether runit is slower than s6 remains to be properly seen. What was proffered isn’t a valid reason or comparison. It’s a supposition and opinion that should’ve been left unsaid, to be honest as it’s iffy at best with how it was presented (Not a foregone and it was stated as such…) and is wading off into the same territory as the claims and arguments made about systemd- which we all know are garbage. We don’t want to do this, gang.l


  11. By the way, s6/66 were installed and used on a Kiss installation as well, you may look it up at It worked first try.

    The comparisons with systemd have always been on an irrational base, by irrational fans of systemd. Systemd is rarely criticised as an init or a service supervisor/manager, but that it extends its reach on all other aspects where an init/svc.supervisor shouldn’t. None of the other systems mentioned here can be accused of doing such a thing as systemd does. So let’s not carry the irrationality here to defend any particular system not for what it is but because it is “criticism” that brought the systemd domination. This is a major fallacy.


  12. [Just here duplicating a post I left somewhere else.]

    Hi yee all,

    I’m not in charge of KISS Linux, just interested by this distro –which I have not installed by the way.

    The link to KISS Linux is broken. The former site seems abandoned. Please now follow:

    It seems that the same Dylan has also conceived neofetch & sowm



If your comment is considered off-topic a new topic will be created with your comment to continue a different discussion. This community is based on open and free communication, meaning we must all respect all in minimizing the exercise of freedom to disrupt such communication. Feel free to post what you think but keep in mind the subject matter discussed. It is just as easy to start a new topic as it is to dilute the content of an existing discussion.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.