The problem with old computers isn’t that they’re slow but fail randomly so they don’t need “smaller” Linux, they need more resiliency that can work with random RAM erros, corrupt disks, absurd CPU instruction failures.
The real issue is that old hardware use a lot of electrical power. You can get a small Single Board Computer with at least as much computing power as those but using 20 to 30 time less electrical power, and fitting in the palm of your hand.
Distros like these help troubleshoot boxes that are old/slow but also not used as computers in the traditional sense. For example network boxes, NAS, video recording boxes etc that can't run the latest LTS ubuntu well but can boot a distro like DSL. getting a vga out on these things with a fast to boot distro helps you fix things like corrupt drives, bad partitioning, bad boot loaders etc which needs a few terminal commands and a distro that boots up quickly.
It once took ubuntu 18.04 30 minutes to boot on an old dual core intel network box once. I switched to Xubuntu and it was about 5 minutes. imagine having to do multiple reboots.
The GUI is usually the problem. I have booted Xubuntu and it's still slow. Slow systems with older GPUs simply can't keep up with newer desktops. Most of the times I need a terminal but a lightweight desktop can help if I quickly want to open a browser and search something so that I can copy past more complicated commands.
I recently used it to boot a ~1996 Compaq Presario from CD-Rom to image the hard-drive to a USB stick before wiping it for my retro-computer fun :)
It's kind of sad to hear "adult" people claim in all seriousness that it's reasonable that a kernel alone spends more memory than the minimum requirement for running Windows 95, the operating system with kernel, drivers, a graphical user interface and even a few graphical user-space applications.
I got this insight from a previous thread : you can run linux with gui on the same specs as win 95 fine if your display resolution is 640x480. The framebuffer size is the issue
I mean why is that a problem? Win95 engineering reflects the hardware of the time, the same way today's software engineering reflects the hardware of our time. There's no ideal here, there's no "this is correct," etc its all constantly changing.
This is like car guys today bemoaning the simpler carburetor age or the car guys before them bemoaning the model T age of simplicity. Its silly.
There will never be a scenario where you need all this lightweight stuff outside of extreme edge cases, and there's SO MUCH lightweight stuff its not even a worry.
Also its funny you should mention win95 because I suspect that reflects your age, but a lot of people here are from the dos/first mac/win 2.0 age, and for that crowd win95 was the horrible resource pig and complexity nightmare. Tech press and nerd culture back then was incredibly anti-95 for 'dumbing it all down' and 'being slow' but now its seen as the gold standard of 'proper computing.' So its all relative.
The way I see hardware and tech is that we are forced to ride a train. It makes stops but it cannot stop. It will always go to the next stop. Wanting to stay at a certain stop doesn't make sense and as in fact counter-productive. I wont go into this, but linux on the desktop could have been a bigger contender if the linux crowd and companies were willing to break a lot of things and 'start over' to be more competitive with mac or windows, which at he time did break a lot of things and did 'start over' to a certain degree.
The various implementations of linux desktop always came off clunky and tied to unix-culture conventions which dont really fit the desktop model, which wasn't really appealing for a lot of people, and a lot of that was based on nostalgia and this sort of idealizing old interfaces and concepts. I love kde but its definitely not remotely as appealing as win11 or macos gui and ease of use.
In other words, when nostalgia isn't pushed back upon, we get worse products. I see so much unquestionable nostalgia in tech spaces, I think its something that hurts open source projects and even many commercial ones.
> There will never be a scenario where you need all this lightweight stuff
I think there are many.
Some examples:
* The fastest code is the code you don't run.
Smaller = faster, and we all want faster. Moore's law is over, Dennard scaling isn't affordable any more, smaller feature sizes are getting absurdly difficult and therefore expensive to fab. So if we want our computers to keep getting faster as we've got used to over the last 40-50 years then the only way to keep delivering that will be to start ruthlessly optimising, shrinking, finding more efficient ways to implement what we've got used to.
Smaller systems are better for performance.
* The smaller the code, the less there is to go wrong.
Smaller doesn't just mean faster, it should mean simpler and cleaner too. Less to go wrong. Easier to debug. Wrappers and VMs and bytecodes and runtimes are bad: they make life easier but they are less efficient and make issues harder to troubleshoot. Part of the Unix philosophy is to embed the KISS principle.
So that's performance and troubleshooting. We aren't done.
* The less you run, the smaller the attack surface.
Smaller code and less code means fewer APIs, fewer interfaces, less points of failure. Look at djb's decades-long policy of offering rewards to people who find holes in qmail or djbdns. Look at OpenBSD. We all need better more secure code. Smaller simpler systems built from fewer layers means more security, less attack surface, less to audit.
Higher performance, and easier troubleshooting, and better security. There's 3 reasons.
Practical examples...
The Atom editor spawned an entire class of app: Electron apps, Javascript on Node, bundled with Chromium. Slack, Discord, VSCode: there are multiple apps used by tens to hundreds of millions of people now. Look at how vast they are. Balena Etcher is a, what, nearly 100 MB download to write an image to USB? Native apps like Rufus do it in a few megabytes. Smaller ones like USBimager do it in hundreds of kilobytes. A dd command in under 100 bytes.
Now some of the people behind Atom wrote Zed.
It's 10% of the size and 10x the speed, in part because it's a native Rust app.
The COSMIC desktop looks like GNOME, works like GNOME Shell, but it's smaller and faster and more customisable because it's native Rust code.
GNOME Shell is Javascript running on an embedded copy of Mozilla's Javascript runtime.
Just like dotcoms wanted to dis-intermediate business, remove middlemen and distributors for faster sales, we could use disintermediation in our software. Fewer runtimes, better smarter compiled languages so we can trap more errors and have faster and safer compiled native code.
Smaller, simpler, cleaner, fewer layers, less abstractions: these are all goods things which are desirable.
Dennis Ritchie and Ken Thompson knew this. That's why Research Unix evolved into Plan 9, which puts way more stuff through the filesystem to remove whole types of API. Everything's in a container all the time, the filesystem abstracts the network and the GUI and more. Under 10% of the syscalls of Linux, the kernel is 5MB of source, and yet it has much of Kubernetes in there.
Then they went further, replaced C too, made a simpler safer language, embedded its runtime right into the kernel, and made binaries CPU-independent, and turned the entire network-aware OS into a runtime to compete with the JVM, so it could run as a browser plugin as well as a bare-metal OS. Now we have ubiquitous virtualisation so lean into it: separate domains. If your user-facing OS only runs in a VM then it doesn't need a filesystem or hardware drivers, because it won't see hardware, only virtualised facilities, so rip all that stuff out. Your container host doesn't need to have a console or manage disks.
This is what we should be doing. This is what we need to do. Hack away at the code complexity. Don't add functionality, remove it. Simplify it. Enfo...
I wonder if it would be possible to have gone directly to Zed, without going through Atom first (likewise, Plan 9 would never have been the first iteration of a Unix-like OS). "Rewrite it in Rust" makes a lot of sense if you have a working system that you want to rewrite, but maybe there's a reason that "rewrite it in Rust" is a meme and "write it in Rust" isn't. If you just want to move fast, put things up on the screen for people to interact with, and figure out how you want your system to work, dynamic languages with bytecode VMs and GC will get you there faster and will enable more people to contribute. Once the idea has matured, you can replace the inefficient implementation with one that is 10% of the size and 10x the speed. Adding lots of features and then pruning out the ones that turn out to be useless may also be easier than guessing the exact right feature set a priori.
if you can compile the kernel though, there is no reason that W95 should be any smaller than your specifically compiled kernel - in fact it should be much bigger
You can run an entire, productive modern Linux, (minus a modern browser*) on 128MB of RAM and one slow core. If you push lower, you start running into issues. I would recommend having around 200MB or so of swap for sudden spikes of memory usage. An aggressive userspace OOM killer may make life easier.
On Linux, if you just run SDDM that launches xfce, you will quickly OOM the system, because SDDM will stay in memory. The same goes for most desktop managers. So the real way is to just `startx` your desktop environment directly and use console login.
i3 is the best call for usability/a modern feeling, with extremely low memory usage. The reasoning is that, if you're used to sway or i3, this will feel like home, and it has all the features you need to be productive. Anything else will eat more RAM, from what I've tried. It also feels really fast, even if your GPU is struggling, because there are no animations and very little movement.
I would personally recommend Alpine, as it really comes with nothing extra. You can then install a desktop environment manually (or use their setup-desktop script if you have plenty of RAM and storage). TinyCore is a bit too wild to do modern computing on; the paradigms are just too outdated, the installation is a bit of a pain, and the installer would OOM on the same system where I can run my entire i3 alpine setup.
DSL seems cool, I haven't tried it; I just wanted to share my experience.
You can try all of this by setting up a qemu VM. Be aware that you will need more RAM just for the BIOS, so maybe if you configure 210MB, youll end up with around 128 usable, or so. Your OS will report how much is usable, accurately.
You can then set CPU cores with usage limits, limit HDD speeds to 2000's HDD speeds (so that your swap isnt fast), and so on. Its a fun exercise to try to develop software on such a limited machine, and its fun to see which software launches and which doesn't.
*: the browser is an issue. Firefox is the preferrable option, but wouldn't launch on so little RAM. NetSurf or elinks/lynx etc. is the way to go, but websites that rely on JS to work, like any and all Rust documentation sites, will be completely unusable.
Don't forget the links2 classic web browser! (still missing the <video> and <audio> element support on x11/wayland though).
Server side rendering will collect(steal) personal info, it is a no go. The only right solution is online services to provide a web site on the side of a whatng cartel web app, if the online service does fit ofc. No other way, and hardcore regulation is very probably required.
"Back in the day" people were running HP technical workstations, with X11 with CDE, with 128MB RAM, on Pentium-II equivalent speed CPUs - and they liked it!
Another distro worth noting here is EasyOS, a current project by Puppy Linux creator Barry Kauler: https://easyos.org/
I remember having tested it, but can't remember what it was like :) -- at least it didn't make me switch from Tiny Core Linux, which I've used extensively. From a superficial, distro-hopper view, DSL, Puppy, EasyOS and Tiny Core all feel quite similar, I guess.
As a side note, it is interesting to see DSL and TC on the HN front page in two consecutive days of 2025. Both are very old projects; I wonder what's the impulse behind this current interest.
Took me a moment to understand what "DSL" relates to in this context. I thought of "dynamic scripting language" at first and was confused. But here it just means "Damn Small Linux".
Back in the early 00's I used DSL in university as it has an x86 emulator included. I could plug the USB in and run Linux under Windows. Kept that on a "huge" 512MB thumb drive that cost over $100. Still have that drive and it still works.
Damn Small Linux was one of the first distros I tried out as teenager with a LiveCD. It's sad it fell out due to his lead being rather an asshole with his contributors and overall incompetent.
BTW, one of former core contributors went to make their own distro called TinyCoreLinux.
I didn't know it was revived, had some fun with it back in the day. I'm curious to see how it compares against Alpine which is also very compact because of musl.
The other day, another os - tiny core Linux - was posted. Today this, I really wonder what is the trade off? Day to day use, security, and something else surely will be missing?
I used to use this on a CD-ROM, for SSH-ing into my personal server to check email (from work or SO's place), when I didn't have a laptop or handheld with me. USB flash drives often didn't boot by default on PC hardware, but CD-ROMs did.
Later, I made an immutable "Swiss Army knife" USB stick distro called LilDeb, based on Debian Live. And literally carried it on a Swiss Army pocketknife on my keychain. LilDeb was implemented in two big shell scripts, which you can still see at: https://www.neilvandyke.org/lildeb/
I remember backpacking in Asia I kept a DSL bootable USB with me. When I'd visit a net cafe I'd simply boot into my DSL environment and bypass the entire windows PC (which was full of spyware and password stealing programs).
It worked great back then, I'm sure it works even better now.
35 comments
[ 2.8 ms ] story [ 68.6 ms ] threadhttps://web.archive.org/web/20100520020401/http://therealedw...
The size was a 90s problem.
It once took ubuntu 18.04 30 minutes to boot on an old dual core intel network box once. I switched to Xubuntu and it was about 5 minutes. imagine having to do multiple reboots.
It fits into under 700 MB and runs in well under 100 MB of RAM. The default Ubuntu image is about 6 GB now and takes a gig of RAM.
Have you not tried it? I have:
https://www.theregister.com/2024/02/14/damn_small_linux_retu...
It's kind of sad to hear "adult" people claim in all seriousness that it's reasonable that a kernel alone spends more memory than the minimum requirement for running Windows 95, the operating system with kernel, drivers, a graphical user interface and even a few graphical user-space applications.
This is like car guys today bemoaning the simpler carburetor age or the car guys before them bemoaning the model T age of simplicity. Its silly.
There will never be a scenario where you need all this lightweight stuff outside of extreme edge cases, and there's SO MUCH lightweight stuff its not even a worry.
Also its funny you should mention win95 because I suspect that reflects your age, but a lot of people here are from the dos/first mac/win 2.0 age, and for that crowd win95 was the horrible resource pig and complexity nightmare. Tech press and nerd culture back then was incredibly anti-95 for 'dumbing it all down' and 'being slow' but now its seen as the gold standard of 'proper computing.' So its all relative.
The way I see hardware and tech is that we are forced to ride a train. It makes stops but it cannot stop. It will always go to the next stop. Wanting to stay at a certain stop doesn't make sense and as in fact counter-productive. I wont go into this, but linux on the desktop could have been a bigger contender if the linux crowd and companies were willing to break a lot of things and 'start over' to be more competitive with mac or windows, which at he time did break a lot of things and did 'start over' to a certain degree.
The various implementations of linux desktop always came off clunky and tied to unix-culture conventions which dont really fit the desktop model, which wasn't really appealing for a lot of people, and a lot of that was based on nostalgia and this sort of idealizing old interfaces and concepts. I love kde but its definitely not remotely as appealing as win11 or macos gui and ease of use.
In other words, when nostalgia isn't pushed back upon, we get worse products. I see so much unquestionable nostalgia in tech spaces, I think its something that hurts open source projects and even many commercial ones.
I think there are many.
Some examples:
* The fastest code is the code you don't run.
Smaller = faster, and we all want faster. Moore's law is over, Dennard scaling isn't affordable any more, smaller feature sizes are getting absurdly difficult and therefore expensive to fab. So if we want our computers to keep getting faster as we've got used to over the last 40-50 years then the only way to keep delivering that will be to start ruthlessly optimising, shrinking, finding more efficient ways to implement what we've got used to.
Smaller systems are better for performance.
* The smaller the code, the less there is to go wrong.
Smaller doesn't just mean faster, it should mean simpler and cleaner too. Less to go wrong. Easier to debug. Wrappers and VMs and bytecodes and runtimes are bad: they make life easier but they are less efficient and make issues harder to troubleshoot. Part of the Unix philosophy is to embed the KISS principle.
So that's performance and troubleshooting. We aren't done.
* The less you run, the smaller the attack surface.
Smaller code and less code means fewer APIs, fewer interfaces, less points of failure. Look at djb's decades-long policy of offering rewards to people who find holes in qmail or djbdns. Look at OpenBSD. We all need better more secure code. Smaller simpler systems built from fewer layers means more security, less attack surface, less to audit.
Higher performance, and easier troubleshooting, and better security. There's 3 reasons.
Practical examples...
The Atom editor spawned an entire class of app: Electron apps, Javascript on Node, bundled with Chromium. Slack, Discord, VSCode: there are multiple apps used by tens to hundreds of millions of people now. Look at how vast they are. Balena Etcher is a, what, nearly 100 MB download to write an image to USB? Native apps like Rufus do it in a few megabytes. Smaller ones like USBimager do it in hundreds of kilobytes. A dd command in under 100 bytes.
Now some of the people behind Atom wrote Zed.
It's 10% of the size and 10x the speed, in part because it's a native Rust app.
The COSMIC desktop looks like GNOME, works like GNOME Shell, but it's smaller and faster and more customisable because it's native Rust code.
GNOME Shell is Javascript running on an embedded copy of Mozilla's Javascript runtime.
Just like dotcoms wanted to dis-intermediate business, remove middlemen and distributors for faster sales, we could use disintermediation in our software. Fewer runtimes, better smarter compiled languages so we can trap more errors and have faster and safer compiled native code.
Smaller, simpler, cleaner, fewer layers, less abstractions: these are all goods things which are desirable.
Dennis Ritchie and Ken Thompson knew this. That's why Research Unix evolved into Plan 9, which puts way more stuff through the filesystem to remove whole types of API. Everything's in a container all the time, the filesystem abstracts the network and the GUI and more. Under 10% of the syscalls of Linux, the kernel is 5MB of source, and yet it has much of Kubernetes in there.
Then they went further, replaced C too, made a simpler safer language, embedded its runtime right into the kernel, and made binaries CPU-independent, and turned the entire network-aware OS into a runtime to compete with the JVM, so it could run as a browser plugin as well as a bare-metal OS. Now we have ubiquitous virtualisation so lean into it: separate domains. If your user-facing OS only runs in a VM then it doesn't need a filesystem or hardware drivers, because it won't see hardware, only virtualised facilities, so rip all that stuff out. Your container host doesn't need to have a console or manage disks.
This is what we should be doing. This is what we need to do. Hack away at the code complexity. Don't add functionality, remove it. Simplify it. Enfo...
however this is of course easier said than done
On Linux, if you just run SDDM that launches xfce, you will quickly OOM the system, because SDDM will stay in memory. The same goes for most desktop managers. So the real way is to just `startx` your desktop environment directly and use console login.
i3 is the best call for usability/a modern feeling, with extremely low memory usage. The reasoning is that, if you're used to sway or i3, this will feel like home, and it has all the features you need to be productive. Anything else will eat more RAM, from what I've tried. It also feels really fast, even if your GPU is struggling, because there are no animations and very little movement.
I would personally recommend Alpine, as it really comes with nothing extra. You can then install a desktop environment manually (or use their setup-desktop script if you have plenty of RAM and storage). TinyCore is a bit too wild to do modern computing on; the paradigms are just too outdated, the installation is a bit of a pain, and the installer would OOM on the same system where I can run my entire i3 alpine setup.
DSL seems cool, I haven't tried it; I just wanted to share my experience.
You can try all of this by setting up a qemu VM. Be aware that you will need more RAM just for the BIOS, so maybe if you configure 210MB, youll end up with around 128 usable, or so. Your OS will report how much is usable, accurately.
You can then set CPU cores with usage limits, limit HDD speeds to 2000's HDD speeds (so that your swap isnt fast), and so on. Its a fun exercise to try to develop software on such a limited machine, and its fun to see which software launches and which doesn't.
*: the browser is an issue. Firefox is the preferrable option, but wouldn't launch on so little RAM. NetSurf or elinks/lynx etc. is the way to go, but websites that rely on JS to work, like any and all Rust documentation sites, will be completely unusable.
Server side rendering will collect(steal) personal info, it is a no go. The only right solution is online services to provide a web site on the side of a whatng cartel web app, if the online service does fit ofc. No other way, and hardcore regulation is very probably required.
"Back in the day" people were running HP technical workstations, with X11 with CDE, with 128MB RAM, on Pentium-II equivalent speed CPUs - and they liked it!
Cookies setup for HN:
- Use ZRAM: - Tiling sucks on small resolutions. Use CWM or IceWM.- XTerm it's very small and works fine. I can posts Xresorces there.
I remember having tested it, but can't remember what it was like :) -- at least it didn't make me switch from Tiny Core Linux, which I've used extensively. From a superficial, distro-hopper view, DSL, Puppy, EasyOS and Tiny Core all feel quite similar, I guess.
As a side note, it is interesting to see DSL and TC on the HN front page in two consecutive days of 2025. Both are very old projects; I wonder what's the impulse behind this current interest.
There are also some charities that ship old PCs to Africa, install a small Linux distro on them, e.g.:
https://www.computers4charity.org/computers-for-africa
https://worldcomputerexchange.org/
When I lived in London I helped clients donate a lot of kit to ComputerAid International:
https://www.computeraid.org/
And what's now Computers4Charity:
https://www.computers4charity.org/
BTW, one of former core contributors went to make their own distro called TinyCoreLinux.
Later, I made an immutable "Swiss Army knife" USB stick distro called LilDeb, based on Debian Live. And literally carried it on a Swiss Army pocketknife on my keychain. LilDeb was implemented in two big shell scripts, which you can still see at: https://www.neilvandyke.org/lildeb/
It worked great back then, I'm sure it works even better now.