Dusk OS | Hacker News Reader

Dusk OS

duskos.org

183 pointsGTP112 comments

>Its primary purpose is to be maximally useful during the first stage of civilizational collapse, that is, when we can't produce modern computers anymore but that there's still many modern computers around.

Any event or series of events that removes mankind's ability to produce modern computers is a global extinction-level event and rather than dicking around with computers one should really be considering suicide to avoid a slow, painful, inevitable death in a hostile world surrounded by misery.

People act like computers are complicated. They are but they also aren't.

Any moderately-sized US state university can (and some have) build one from scratch: as in from fucking sand to "Shall we play a game?", all in one go.

The state university nearest to me has a complete nanofab that can make-- and package!-- ICs (somewhere around 14nm-ish), a different lab that can make wafers from scratch, a chemistry department where undergrads could make the plastics, and all of the software guys you can shake a stick at.

The loss of the ability to make many things, including computers but also other more important things like the industrial process for making ammonia, globally, simultaneously, is the end of humanity.

The knowledge and ability is so widely globally distributed that taking it all out is death.

Do not mistake the centralization of consumer goods assembly with the centralization of the knowledge needed to assemble consumer goods.

Is this OS just for the brief period of time between the loss and the ultimate end? To like, play some rounds of solitaire while awaiting the inevitable?

>Any moderately-sized US state university can (and some have) build one from scratch: as in from fucking sand to "Shall we play a game?", all in one go.

I am quite surprised to hear that. Really?

Yup!

Here's one in Utah: https://www.nanofab.utah.edu/

Here's one in Pennsylvania: https://www.mri.psu.edu/nanofabrication-lab

Here's one in California: https://nanofab.ucsb.edu/

Those aren't starting from sand though, so they do need some external fancy tech to survive the apocalypse - to produce their wafers.

It won't be small, but you can create PCBs pretty easily with chemical etching. I think making the silicon wafers is probably harder.

PCBs are something people can do in their garage. Integrated circuit fabrication is definitely something that requires a much more sophisticated lab such as a top university lab.

Depends on the state, but for my alma mater (Georgia Tech) I'm pretty confident the answer is yes.

Georgia Tech has one of the best fabrication programs in the world.

I just had a look and I am not sure they are making silicon wafers. I would be surprised, there's not really any research in it, and it's costly to say the least.

Yeah, I find that rather surprising that a university lab could fabricate a practical 14 nm chip (more than just some demonstration gates), given the billions of dollars that countries have to invest to set up even 65 nm fabs.

To your point - we do have some chokepoints in supply chains here and there but I often wonder how much of that is due to the learning curve. As in, things are centralized because just how we as a civilization learned to build things in our particular historical context - if the context changes and the knowledge and resources are there to rebuild in a different way, we can.

Relatedly - a lot of things seem intrinsically capital intensive, but how much of that is due to the fact that we had large pools of capital when we learned to do those things?

Perhaps I'm ignorant, but I wonder why I haven't seen this aspect explored in post-apocalyptic media more. I mean, beyond the obvious fact that if we can "keep making computers" all of a sudden the apocalypse doesn't sound quite that bad, which would somewhat lower the stakes.

> Any event or series of events that removes mankind's ability to produce modern computers is a global extinction-level event and rather than dicking around with computers one should really be considering suicide to avoid a slow, painful, inevitable death in a hostile world surrounded by misery.

Hmmm... At a very high altitude extinction-level events have sufficiently sharp edges. But, as we get closer to things it becomes fuzzy. For examples: The Black Death (1347-1351), The Spanish Flu (1918-1919), The Great Chinese Famine (1959-1961), The Haitian Revolution (1791-1804), The Irish Potato Famine (1845-1852), each one of these events were extinction-level events for those in the middle of them. What year would you have selected to self terminate? Remember you would not have hindsight or knowledge that the events are temporary.

Those were all local events.

The only thing removing mankind's ability to make computers is a large (Chicxulub-ish-sized) meteor strike or andromeda-strain-like sci-fi story made real.

Luckily, I'll be killed by tidal waves if a meteor hits the Atlantic and I'll be killed in the firestorm if it hits land anywhere on Earth.

If it hits the Pacific or Indian oceans, it depends on the size. If large enough, I'll shoot myself in the head to avoid starvation after playing a few rounds of solitare.

> Those were all local events

In a sufficiently catastrophic event (major loss of infrastructure), one wouldn't be able to tell if the event is local or global. To the people suffering from the Black Death, the event must have felt global as everyone they knew was experiencing it.

> Those were all local events.

> The only thing removing mankind's ability to make computers is a large (Chicxulub-ish-sized) meteor strike or andromeda-strain-like sci-fi story made real.

One of the possible causes of the Bronze Age Collapse (which btw was very non-local) is a prolonged drought that apparently lasted for decades. It's wasn't severe by itself but long enough for the fertile soil to keep shrinking year after year, decreasing the size of sustainable population.

I dwell on this sometimes and I think that probably we underestimate how fragile our food production system is. Feeding 8 billion people is not an easy endeavor. Multi-year global food reserves are not only non-existent but maybe even non-possible at this scale. A global event that would somehow disrupt our ability to produce food for several years in a row will make the hell break loose. I'm not confident that humanity couldn't plunge in a couple of dark age centuries as a result, all bets are off in a truly global famine.

Then the curious question is how likely is such an event? Chicxulub -- yes, but these are extremely rare, once in millions of years. The Volcanic Winters however occur every few centuries, the "Year Without a Summer" was 1816. Could we have "3 years without a summer" at some point? And not just in the northern hemisphere but globally? I don't know, maybe.

After all, the way I think of it is this: modern food production at scale is sun + water + fertilizers. Production of fertilizers seem distributed enough to be resilient at a global scale. Water at the Earth scale is sun again -- as long as it's shining it will rain somewhere, even if the distribution shifts (with dramatic effects for sure, but not a completely desperate situation). But then the sun does seem to be the single point of failure. If there's dust or ash or something else shielding the sun then it is in fact desperate.

(disclaimer - I know very little about this, it's just common sense and theorycrafting)

> Multi-year global food reserves are not only non-existent but maybe even non-possible at this scale.

I agree with the sentiment, but I will risk a hypothesis that in "western" world we have enough canned and highly processed food (the one that lies on the shelf of your grocery store and has best before set 2-3 years ahead) to survive "covid but for crops" event. There are couple of reasons I feel safe about this. We waste massive amout of food on a daily basis. We have a huge variety of food available, and not everything has to be affected. We already rely on global supply chains on a daily basis and for whatever reason while living in EU I'm eating garlic imported as far away as China. And then there's coffee, bananas, etc.

Sure, covid has shown that a massive shift will have a dramatic short term impact. Somehow Shkreli was sent to prison after raising price of a drug by 5000%, but when every pharmacy in my country did exactly the same with face masks everyone was quite ok with that, as long as they could get one. And after the initial shock we got back to seamingly normal life.

If anything, I'm scared about the water.

But all in all, when the shit really hits the fan, I envision scenario similar to McCarthy's `The Road` or for lack of better example `The Walking Dead`. People will be more busy trying to secure food and shelter rather than figuring out how to boot some obscure OS on a defunct computer.

I don't think COVID is a good benchmark. The whole matter was relatively minor, claiming lives of 0.09% of earth population (by WHO estimates). We can surely survive "COVID for plants" with little damage but that's not at all the scenario I have in mind. I'll try to clarify on specific bullet points.

> I will risk a hypothesis that in "western" world we have enough canned and highly processed food (the one that lies on the shelf of your grocery store and has best before set 2-3 years ahead) to survive "covid but for crops" event.

The "western world" is definitely more prepared but if you want to rely on that during global and desperate famine you'll need to have the guts (and the heart) to machine gun hundreds of millions of refuges rushing into the western world to survive. Or nuke them before they depart.

> We waste massive amout of food on a daily basis. We have a huge variety of food available, and not everything has to be affected.

Absolutely all the food we have is based on the ability of plants to absorb sunlight. Exception would be an electric light greenhouse powered by the nuclear plant. Or by the coal/oil (that's some sunlight we can actually store). Take away the sunlight and the whole food variety goes down with it.

> And after the initial shock we got back to seamingly normal life.

Try to imagine a desperate famine: your kids have nothing to eat and will die in a matter of weeks unless you do something. There's no reason to believe the situation will improve. In such times killing your neighbor who has some food left keep your kids alive a bit longer is not an unusual decision. Someone has to die anyway, that is so.

Now imagine that at a global scale, famine is everywhere.

Not OP, but I somehow share his sentiment and I'd say that out of your list any would do.

Terrible things are happening right now, and people find their way out, for instance the war in Ukraine. But then again, they have visa cards, they can book travel and accomodation online and flee from the country. A lot of jews have fled from Europe during and before WW2 when things were not so much connected, but still, it was modern day, people had access to press, radio, telegrams and telephones.

But if you live in a rural village and everyone you knew died to a famine that sounds like something you might not want to carry any further.

Yeah we’re not likely to ever face such a situation and still even need that many computers. It’s an interesting project though in the sense that someone has a 32 bit OS running on every ISA out there (or even aims to).

Interesting but not practical. CPUs are either 64 bit and have memory management hardware or 32 bit and don’t. This dictates whether you have lots of addressable RAM or not, and changes what an OS is for that CPU

It is also interesting in the fact that it is extremely tiny in terms of lines of code, and the approach they took to minimize its codebase is original.

To give some color, here: https://blog.sentinel-team.org/p/how-likely-are-various-prec... is a list of a few things that could kill over 1M people in one year. Maybe your statement is not the case for solar storms in particular.

The probability estimation of "simulation shutdown" seems a bit high. It's just as speculative as alien invasions or vacuum decay.

Speculative rationalist dreck. Entertaining nonetheless

We just experienced 10x that 5 years ago. It did damage supply chains but it didn’t result in collapse.

Wow... there are probably several ways we could lose the ability to fab semiconductors. We can look back through history at other periods where they "forgot" how to do various things. Brunelleschi's dome on the Florence Cathedral was the first major dome built in Europe in over a thousand years since the fall of the Western Roman Empire. They lost the knowledge of dome building. Should they have considered doing themselves in because they found themselves in the Dark Ages? We might forget how to do various very involved technical things like fab'ing chips, or we might lose the knowledge/and or ability to even build a fab, for example (just consider the supply-chain required to build a fab and consider how fragile some of those chains are). Yeah, it will mean that something really bad has likely happened, but that doesn't mean that it's the end and we should just go and kill ourselves. It will mean that there's been some kind of discontinuity, but it doesn't necessarily mean the end of humanity. People will still be able to grow crops, hunt, fish, etc (given a reasonable climate remains in enough places). Likely the population under these conditions will be much smaller but it doesn't mean that we can never recover as a species - sure, it might take 1000 years. That's ok, we go on just like ancestors before us who endured great hardship.

The world is more than Europe.

There are many domes that were built in the Middle East and Asia while Europe was trying to figure out how to fund expensive vanity projects (large domes).

Computers aren't vanity projects.

> Computers aren't vanity projects.

Nor are they a requirement for survival. We didn't have computers until about the 1930s and somehow we survived as a species.

> The world is more than Europe.

Yes, very true. But under the conditions of some kind of serious disaster such as we're discussing we wouldn't have visibility into what's going on on the other side of the world either, just as they didn't. The Dark Ages in Europe weren't dark in places like China, but that didn't help you if you were living in Europe. (and even in Europe the "darkness" wasn't evenly distributed)

There are no Dark Ages. People in the Byzantine Empire (capitale in Europe) and the Abbasid Caliphate very much knew how to build domes and where actually busy advancing the state of the art.

The fact that the Franks were more interested in warring against each others than building great things is no evidence of a dark age. They started building again once peace came back and didn’t restart from where they left off but from the new state of the art as translations started pouring over from the Arabic empires.

The idea that humanity somehow went back is a fantasy pushed forward by catholic commentators.

The 'dark ages' weren't even the dark ages in Europe. The idea was invented by guys in the Renaissance who pined for the Greco-Roman world and pretended everything since then was a slide downhill. Anyone who seriously trots out this trope as 'proof' of anything should be taken about as seriously as a flat-earther.

> The idea that humanity somehow went back is a fantasy pushed forward by catholic commentators.

A tangential aside: Wouldn't that make the catholics look bad and as such why would they spread it? ("We had all this great tech before the catholic/christians came along and ruined everything and now we don't have nice things" seems like another way to look at it)

Because through racist-enough eyes, Arabic discoveries or knowledge don't exist until they are whitewashed. That's the whole pretense, that there was some Dark Age that did something to the whole world. For values of "whole" meaning "the parts we care about".

> "We had all this great tech before the catholic/christians came along and ruined everything and now we don't have nice things" seems like another way to look at it

The way it was framed (and often still is with a dash of 19th century nationalism added) is:

The unwashed barbarian from the east came pouring tearing apart the beautiful christian roman empire before one of their chief saw the light and converted. Guided by faith, the franks unified once again the warring provinces which sadly were broken apart again through inheritences - but such is the law - before their god annointed kings heroically pushed away the unfaithful from the south. The properous Italian cities, whose families gave us more Pope than any others, rediscovered the brightness of the Antiquity especially the best of them, Aristotle, who turns out to be the easiest to reframe in a way compatible with catholicism, propelling the world into the Age of Enlightnment which had to be spread out to the furthest shores.

> Brunelleschi's dome on the Florence Cathedral was the first major dome built in Europe in over a thousand years since the fall of the Western Roman Empire.

I don't think it necessarily means they forgot how to do it, though. Or does it? We too stopped building domes made of big stones, it doesn't mean we don't know how to do it.

For whatever reason they weren't building domes anymore in Europe where they had been (the fall of the Western Roman empire sort of changed priorities for a while). And after a generation of not building domes the artisans (in Europe) lost the ability to do it.

There's a much closer example: The US spent the 60s developing the capability to land humans on the moon. And they were successful. They did it about, what, 7 or 8 times? But then they stopped doing it. And now some 60 years later they're having a hard time doing it again (see the woes of NASA's Artemis and Boeing's Starliner). Imagine if the pause wasn't 60 years, but several hundred years.

That's a really good example.

I only want to add that's not something that we have a hard time doing because it suddenly became hard, but because it came out of fashion.

We had a really good time thanks to cold war and flexing between both sides, but now it's a just a enormous investment with no return. It's really sad to see NASA running on a shoe string budget, and Ingenuity being nothing more than a glorified student project that happened at the right time in the right place.

Some people would survive.

That was a hard, and good read. But remember Linux is Love. Linux is Life.

If the end of the world happens I'm going down making and doing dope stuff. I'm not going outside and mingling with people. That's how you get a fucking spear thrown at you or mugged by some newfound MS14 gang swinging hockey sticks and shit.

I'm going to forage for some fruits and vegetables, all that good stuff since I'm too much of a bitch to kill an animal; those things are cute.

I'm going to load up my laptops on solar power/battery, hook in to some internet somewhere and live a normal life.

I don't know what Dusk OS is about but the making computers from fucking sand, now that's dank.

I'll install Desktop Linux on it and we can all start the next era of civilization; this time Microsoft and Apple aint selling us out

catches spear to the chest Shit

> Any event or series of events that removes mankind's ability to produce modern computers is a global extinction-level event

Semiconductors have been around for, what, 70 years? The idea that humans couldn’t survive without modern tech is ludicrous.

Sure, there will be fewer humans, but extincting humans completely is likely to be incredibly difficult without an atmosphere-on-fire scale event.

I think the stated claim is that any event that takes out semiconductors takes out humanity, not that the lack of semiconductors itself does.

The point isn't that humans couldn't survive without computers, but that any even that managed to stop humans globally from manufacturing computers would be such a massively disruptive event in much more important areas than computing.

> massively disruptive event

That’s not the same as what I was replying to: “a global extinction-level event”

The difference is pretty substantial.

> The loss of the ability to make many things, including computers but also other more important things like the industrial process for making ammonia, globally, simultaneously, is the end of humanity.

The end of us sedentary people. No more ammonia by itself won’t affect hunter-gatherers and uncontacted peoples. But if we’re talking about not the end of these things in itself but an event like a nuclear winter: sure!

We’ve been through population bottlenecks in the distant past. Again, if we (reading this) die out not due to very severe climate change or nuclear winter, other kinds of humans could live on. Maybe.

By the way I think old-school Preppers are funny.[1] What’s the point of prepping with a bunker and canned goods? Truly. You don’t have enough time to train for the skills, or to hoard enough stuff, to survive in a post-civilizational state. You’re just prepping for surviving a few months at best, that’s it.

[1] This OS guy is probably more reflected.

Most proper preppers have years of food not months.

I said they would survive for months.

I know. I didn't misread it.

I didn’t misspeak.

> to avoid a slow, painful, inevitable death in a hostile world surrounded by misery.

Well aren't you a bundle of joy, recommending death to people in bad situations.

A life of misery isn't all that bad. Much worse could happen.

This reminds me of A Man For All Seasons:

Thomas More: "At the worst we could be beggars and still be keep company and be merry together."

Lady Alice: "Aha, merry."

Thomas More: "Aye, merry!"

[1] https://www.youtube.com/watch?v=sEa6XRZa9Vg&t=5459s

That advice sounds to me like it’s more for “an entire world where all people are in a bad situation from which they will never escape and neither will their next of kin” but if that comes to pass, the lighthearted talk of embracing the misery will probably end around the time the starvation begins.

Eschatology is always a controversial subject. Mostly pointless to discuss in such a diverse environment.

They can make their own ingots and ultra pure chemicals, and turbopumps and ion implantation machines and spin coaters CVD machines and and the HEPA filters for the cleanroom, and the remote servers that control the licensing dongle for the robot arm that movies the thingie from "A" to "B"?

Sand to "hello world" is a tall order. There is a pyramid of industry that supports the entire thing. Even Sand -> 99.9999 % Si (purification & Czochralski crystal growth) needs multi‑story furnaces, vacuum pulling stations, 10 MW of steady power, and months of process engineering. All of have huge dependency chains.

Even modern "mild" events like supply chain disruptions would be enough to shut down any sort of non-toy level production in short order.

So yeah, a university as it currently stands can make a microchip, relying on all of these dependency chains being in operational order. But I don't doubt that would quickly no longer be the case if you had a hot war or societal collapse.

The exponential growth in processor capabilities relied on a global manufacturing infrastructure from the 1970s until today. To do the same level of progression would require much of that infrastructure.

Assuming someone managed to strategically nuke every major foundry / chemical / machine supplier, but leave everything else the same? Sure, maybe we get back to our present tech in 15-20 years.

Assume it's the result of a societal collapse? No way in hell.

Most catastrophic scenarios I can imagine result in loss of advanced silicon production, but not necessarily all silicon production. More likely than not there would be ways to keep fabs for old well-trodden processes running. In that situation, there’d probably be a temporary shock of no computers followed by new computers being a lot less powerful… it’d be less “no more computers” and more “new computers are comparable to their Pentium 4/PowerPC G5 era counterparts”, in which case operating systems would still be graphical, just a lot more lightweight. Linux and BSDs would probably quickly take over desktop computing due to being able to run reasonably on such systems with minimal legwork.

If we can’t manufacture chips at all, yeah, things will be in enough of a pickle that computers will be the least of our worries.

Dude your existential angst is painfully showing here - are you ok?

Like, what's going on? The pessimism is pretty high. Are you alright?

There's a lot of idiosyncratic verbiage about stuff like collapse of civilization and users vs. operators, but it's not clear to me what makes this more lightweight or hackable than a typical lightweight RTOS or even just baremetal drivers running on a microcontroller.

Or why an "almost C compiler" is better than a C compiler.

I think that having an "almost" C compiler is a trick they use to minimize the size of the codebase.

Yes, this is it. A C compiler in 50 kilobytes of code is pretty innovative.

It's also doesn't lie about the "almost". It's really almost C. It's just that C's stdlib is POSIX centric and doesn't fit Forth well, so it's not implemented as is.

TCC is 100 kB, already exists, and is a real C compiler (+preprocessor+assembler+linker). It can build Linux too!

https://bellard.org/tcc/

100kb binary. I'm talking about source code size. TinyCC's source is 1400kb in size.

Or rather, why is it not much worse than an actual C compiler.

In an apocalyptic event energy would be precious. You'd rather have an optimizing compiler to run longer off your solar-charged car battery or whatever.

I like the roleplay/lore and the tech effort is impressive but I don't think it fits the idea.

> I like the roleplay/lore and the tech effort is impressive but I don't think it fits the idea.

"An optimizing compiler requires at least 500MB of RAM, which can only be looted from level 7 Terminator elites!"

I think many of us feel a lot of existential dread about the future these days, and if hacking on an OS helps the author cope with those feelings and improves their psychology, then I'm all for it. Whatever keeps you sane.

Seems like a wishful thinking about the first stage portion of collapse. You're going to be more interested in running some obsolete version of Windows that controls some proprietary infrastructure hardware, not running calculations of whatever kind or building things from scratch.

But if you an need an OS for which no viruses or exploits have been develop yet, you may be in luck!

Would be funny to finally find an undamaged computer, boot it up and see a ransomware notice.

While it is most likely a very fanciful pet project borne out of post-apo fantasy doomporn, there is one fatal fly in the ointment: In a post apocalyptic situation, who will have THE TIME (nevermind the resources) to dedicate to "operating" one such OS? Apocalypse is generally synonymous with population collapse. Many less humans to interact with means a greatly diminished division of labour and rebuilding civilisation will require first and foremost food production, 14 hours of daily work in the fields, scavenging, hunting, etc....

Taking one worker away from these (tedious) activities would likely be considered an investment that requires significant returns, results, be it only to justify this to other workers who would also like to be sat in front of a computer. Hard to justify having swapped a hard days work for some code on a computer screen.

For that kind of scenario, a lightweight Android rom with some ham-radio driver would probably be more appropriate.

See https://collapseos.org/why.html , under "Aren't there more important things to do than an operating system?"

Depends on the level of collapse, and say you get infrastructure up and running (temporarily, mind you, without manufacturing..) - what then? That's what this is for. Cobbling together the pieces left over.

That's assuming you can find the proprietary software that is needed. In some cases you're better off just cutting the proprietary bits out entirely and hooking straight into the motor controllers... or whatever else is more immediately upstream of the "business end" of the thing you're hacking.

Recent issue of Wired has an interview with the author https://www.wired.com/story/forth-collapse-os-apocalypse-pro...

Tbh I agree with almost every word there, but I can't believe this OS would be of any help in such a situation.

In dusk of our civilization it will not be possible to boot this OS from any modern computer. Those things nowadays don't even have CD. everything is dependent on internet, our data is being disowned by us and put into cloud. In case of broken internet the knowledge is going to be lost.

And how we're going to connect to this site using http? Via dialup?

I think the idea would be to prepare beforehand and have your own copy, alongside your own Wikipedia dump.

You don't need a CD drive to boot an OS. A humble USB stick does the trick. And nearly every computer made in the last 10 years has a USB port and the ability to boot from it

Sometimes it's only usb c. But... We're speaking about dusk of civilization. Pendrive don't live as long as CDs

I have a stack of CD drives. I'll trade you one for food and medications.

What kind of apocalyptic event that wipes out all computer manufacturing capability would let USB drives (and compatible computers) survive? And how would you power that thing?

Let me preface that I do not subscribe to the ideology at all, but the concept of bootstrapping a complex system is intellectually worthwhile and often practical in its use-cases. For example, imagine a tech company with a bunch of datacenters that boot from the network, have redundancy etc, but if you somehow turn the macthines off all at once, no one knows if they can reboot safely (say, service discovery cannot find the machine hosting OS images anymore).

As a practical matter, you can flip the author's mental pessimism and look at the optimistic situation where we conquer other planets, in which case thinking about an optimal bootstrapping path is going to be essential for the society and its industrial needs.

It is also true that some fairly local shocks in human society can cause a cascading failure and doing this analysis fairly regularly can help identify and minimize the cascade.

Collapse OS - https://news.ycombinator.com/item?id=43482705 - March 2025 (199 comments)

Running CollapseOS on an Esp8266 - https://news.ycombinator.com/item?id=38645124 - Dec 2023 (1 comment)

Dusk OS: 32-bit Forth OS. Useful during first stage of civilizational collapse - https://news.ycombinator.com/item?id=36751422 - July 2023 (141 comments)

DuskOS: Successor to CollapseOS - https://news.ycombinator.com/item?id=36688676 - July 2023 (4 comments)

Collapse OS – Why? - https://news.ycombinator.com/item?id=35672677 - April 2023 (1 comment)

Collapse OS: Winter is coming - https://news.ycombinator.com/item?id=33207852 - Oct 2022 (2 comments)

Collapse OS - https://news.ycombinator.com/item?id=31340518 - May 2022 (8 comments)

Collapse OS Status: Completed - https://news.ycombinator.com/item?id=26922146 - April 2021 (2 comments)

Collapse OS – bootstrap post-collapse technology - https://news.ycombinator.com/item?id=25910108 - Jan 2021 (116 comments)

Collapse OS Web Emulators - https://news.ycombinator.com/item?id=24138496 - Aug 2020 (1 comment)

Collapse OS, an OS for When the Unthinkable Happens - https://news.ycombinator.com/item?id=23535720 - June 2020 (2 comments)

Collapse OS - https://news.ycombinator.com/item?id=23453575 - June 2020 (15 comments)

Collapse OS – Why Forth? - https://news.ycombinator.com/item?id=23450287 - June 2020 (166 comments)

Collapse OS – Why? - https://news.ycombinator.com/item?id=22901002 - April 2020 (3 comments)

'Collapse OS' Is an Open Source Operating System for the Post-Apocalypse - https://news.ycombinator.com/item?id=21815588 - Dec 2019 (3 comments)

Collapse OS - https://news.ycombinator.com/item?id=21182628 - Oct 2019 (303 comments)

I feel I can empathize with the ideology, and I want to argue that an idea like this is a big piece of some puzzle.

The puzzle: During the collapse, I think the biggest thing we would lose is our ability to communicate with each other in a reliable way. Like telecommunications and internet and all that. Who and what do we trust?

I imagine that before a collapse like this would happen, people would get a hint to gather as much as they can to help with their survival. Those things would include: this OS, the knowledge to load a sequence of bytes from whatever device is holding the OS to as many CPUs/controllers as possible, strategies to connect any number of arbitrary CPUs to radio devices, some knowledge of public keys and private keys. maybe all neatly organized into a handbook

The OS itself would be responsible for providing as easy an interface for any average joe to generate public/private keys, communicate with other people who have followed their same protocol, and use those public keys to build trust from communications. before the collapse happens, you may even collect a list of public keys you are likely to already trust.

The OS could maybe even have software for building communities of trust or even handling adhoc finances through (don't hate me for this) cryptocurrency.

anyways, this is all based on an assumption that the ability to communicate quickly (and build trust in a decentralized yet controllable way) is the best mitigation we have to a collapse

this answer is slightly influenced by the movie leave the world behind lol

Mandatory question for all OSes - Does it run EMACS or does it run inside of a EMACS?

only Vi, the resource collapse resistant branch /j

VI runs EMACS?! I knew the opposite to be true, but...

;-p

I've seen this before, thought it was super interesting, kudos to the author. Leaving this comment for contrast with some other negative comments.

If you actually want to offer substantive contrast, then I'd suggest elaborating on what you find interesting.

I think that in the event of a such a collapse, there would be two new directions: turning the surviving advanced digital devices into general computers, and building new electronics. Reusing existing common software would be easier than building something from scratch. So, I'd expect that postmarketOS and flavours of Linux for low-power machines meets most needs.

Also, it'd be fantastic if iPhones have a doomsday switch that untethers them from Apple - that'd be the difference between a useless brick and a precious artifact of a bygone era. The post-apocalyptic setting has potential for a game that comes from the perspective of a builder, and goes deeply into civil engineering and IT - build architecture that can withstand the elements, design a water chip, write embedded software for it.

I hope the author doesn't take this the wrong way, but this feels like another Temple OS driven by a different (perhaps milder) kind of insanity.

It's quite technically impressive, but I'm not sure I'd know what to do with it if I were experiencing an apocalypse, in a way that would be more useful to me than an install of Windows XP or Linux or Freedos would be

Temple OS came to my mind, too. At least that one had a UI and games; a much better option if civilization collapses.

It's just such a boutique thing, saying you're prepared to run 32-bit software on hardware during an apocalypse. If CPUs suddenly become scarce, I would have figured a tube-based, relay-based, or Enigma-type machine with specific useful applications would be more practical.

You do realize that your usb charger has a microprocesor that is as capable as computers you used to play on as a child?

Seriously, finding a running 386 today will probably be more challanging than finding a PC capable of running Cyberpunk in 4k 5 years after the fallout.

Like others have mentioned at worst it's another TempleOS or subtle way to point out to your acquaintance that they shouldn't stay with you alone at best.

> years after the fallout.

Ok, but that's not the purpose of the OS. I recommend you to read the Why page:

https://collapseos.org/why.html

Two relevants parts:

> Computers, after a few decades, will break down beyond repair and we won't be able to program microcontrollers any more.

> It's not that modern ICs are more fragile than old ones (maybe they are, after all, more can go wrong with millions of transistors than with thousands), it's that old designs are simpler than modern ones and thus more repairable.

The complete reflexion and arguments are way deeper and interesting through. The most interesting parts is under "Electronics evolved rapidly without any help before. Why can't we pull it again?"

Yeah, I've read that with the voice of Ron Perlman: "FAT, FAT never changes...".

I get that not every vault dweller will be technically apt enough to build a working computer out of a usb charger, but there is plenty of hardware and software that will be laying around in shrines near ghoul settlements.

Every store has POSes that can be easilly repurposed, there's DOS, Windows95 and if your tribe really wants to get an edge during a siege on enemy village you will most likely be able to calculate catapults trajectories on a PIP... I mean old android phone.

An LLM and a modern desktop makes any sufficiently enthusiastic human into an Operator, as the author describes it.

An OS for civilizational collapse is a million times more relevant than an OS dedicated to God.

Churches did a much better job of preserving knowledge through the "Dark Age" than did doomsayer prophets.

I gotta be honest

It didnt start out that way But 14 years later I would say this was also a labor of love and I’m not sure if it was entirely rational:

https://github.com/Qbix

and

https://community.qbix.com

What's this?

I think the idea is:

some computing device is discovered, but the OS is bad/missing/old/not useful etc. Or possibly a novel computer is built from whatever is still around.

The computer itself would still be good if only it had an os.

Dusk OS is easy to port to new arch because it compiles from source on every(?) boot, and only needs a tiny kernel to be ported?

Yes, every time Dusk boots, it boots from a kernel ranging from 2.5 to 4 kilobytes in size, depending on the CPU. The rest is all compiled (very quickly) from source on the fly.

This and more is addressed on the "why.html" page. E.g.:

> (...) you and I are likely to be dead before Collapse OS is actually useful.

or:

> People who are less technical have trouble seeing the point of this project and ask "why can't we just use Linux?". People who have actually lifted Linux' hood don't ask that question because they actually understand the magnitude of complexity involved, but if you haven't, it's a legitimate question to ask.

https://collapseos.org/why.html

Many comments strawman the author then argue with what they imagine he images. Guys just read the 'Why' page [0] and argue with his thesis. I'm not sure to agree with everything but it still is a very interesting and questioning read. Kudos to the author for writing it.

This OS is not for:

- you or your child

- a planet with working computers

- a planet with cheap and abundant 'bootstrap' energy like we had from 1850 to 1960

- a planet where humans really struggle to get food ("you're running away from cannibals", but I think this metaphor is a bad choice after the mad-max refutation)

That page [1] is also a core part of the theory: "What makes collapse inevitable and imminent".

[0] https://collapseos.org/why.html [1] https://collapseos.org/civ.html

> https://git.sr.ht/~vdupras/duskos/tree/master/item/fs/io/kbd...

I don't know forth. Is this a driver? How does it talk to the hardware?

Forth is usually kind of like a macro later on top of Assembly or even machine code in the case of very simple CPUs, although it can be written on top of anything (example: Python). I think a layer on top of C is pretty common these days as the processors keep getting crazy complicated. Another important note is that although there are some conventions, almost everyone rolls their own Forth, so they're all unique in a lot of ways. The old saying is if you've seen one forth, then you've seen one forth. There is an ANS standard, but it isn't used a whole lot to my knowledge.

So I assume it talks to drivers like you would with Assembly? That's just a guess and probably a bad one.

It's just an event loop, watching for status changes in a memory location. Talking to the hardware just occurs when the keyboard changes the values in memory.

The USB host code is in fs/drv/usb: https://git.sr.ht/~vdupras/duskos/tree/master/item/fs/drv/us...

Not that I can read any of it... Forth looks unlike any other language I've ever seen o_O

I think forth is fun, but I can't really take the collapse narrative of this project seriously. IG it's disrespectful, but I kinda view it as a performance art piece with fallout vibes.

Whenever I see something like this I wonder: how is it more valuable than porting freeDOS/busybox to the same environments, with its dearth of existing software and hardware support

That "civilizational collapse" scenario must be straight out of the Fallout game.

In my mind I was reading their page with the voice of Ron Perlman - "FAT, FAT never changes...".

Where can I go to learn how to scavenge for microcontrollers and load/operate an OS like this on them? (Low-key would like to know that I could do that if need be).

That's nice, but it won't replace the trusty abacus.

That's what you need for a real apocalypse.

My solar powered calculator will be fine.

According to the "Linking to this website" section:

  If you want to link to this website, please use http:// links rather than https:// ones. While http:// links are trivially "upgradable" to HTTPS, the opposite is not.

Mods should update the link in this post

Reply on news.ycombinator.com