As always I’m skeptical that this is actually going to be of use in a post-collapse environment, even one where “computers are still available but we can’t build new ones”; I see this as more of a prepper fantasy, albeit one that’s more benign than stockpiling guns in your cabin.
That said, it’s a really interesting idea, literally compiling the C kernel as part of the boot process!
The post didn’t go into detail (that I could find) on how their mini-C differs from ANSI C; I found more info here. Most of the differences seem reasonable, except for the way “&&” and “||” aren’t short-circuiting — that would drive me nuts, and IMO the “real” behavior should be pretty easy to implement; hopefully they’ll add it.
I’m skeptical that this is actually going to be of use in a post-collapse environment
Thought exercise: can you think of any hypothetical software project that would be useful, and that would be foreseeable pre-collapse?
I suspect the answer is there are few to none, because predicting the future to the required degree is just too hard, way harder than writing software (the “fantasy” part of “prepper fantasy” is that the useful skills will be the ones you already have). Maybe the closest you can get is something like the Arctic Code Vault, which might be useful in the event that civilization re-establishes itself.
I do think these sorts of engineering projects are super interesting: safeguarding against civilization collapse is a huge challenge, and I can’t help but wonder what the best approach would be. I just hope the author (or anyone else) doesn’t get sucked in too deep, to the extent that their current quality-of-life suffers.
Most if not all software “post-collapse” is going to be fundamentally useless, because post-collapse I’m concerned about food, water, shelter and coordination. A word processor or an operating system is not going to help me with any of those.
It is nice to re-examine the basis of our computing stacks though. Wish it wasn’t filed so far under the “prepper fantasy” mindset though. It’s a neat thought experiment/premise, much like “if all you had was a machine monitor, how would you build up to something you could actually use.”
It might be more interesting to think about 100 years post-collapse. Getting back to early Victorian technology with knowledge from now but no infrastructure seems feasible. Even without coal, you can create a similar energy density (though not the same volume) with charcoal, though it’s very labour intensive. Once you can smelt iron and copper, you can create wiring and start building electrical generators. Building steam engines is actually harder because the tolerances need to be much higher: you can build a water wheel out of wood and if you can magnetise a couple of lumps of iron then you can generate power (though for a lot of uses you will want a steady voltage and that is a bit more complex, but can be achieved with some damming to regulate the water flow rate).
Getting from there to semiconductors and Isabel software is a lot harder. If you assume some computers survived and you can generate a supply that can power them, what can you do in advance that will minimise the time taken to building a usable fab?
I suspect that the lack of coal is a bigger problem than anything else for building a society that could do this. The Industrial Revolution was possible only because the number of people required to produce food dropped to a small percentage of the population. That, in turn, was possible only because they were replaced with engines powered by cheap energy (steam from coal). It’s possible that wind and water could replace that for small communities but it’s not clear that they’d be large enough to train the specialists required for microelectronics or modern medicine. Stockpiling spare parts for your nearest nuclear power station might be the best bet: if you can run it at 10% peak load, it might run long enough to power your new civilisation to the point that it can build other energy sources.
It might be more interesting to think about 100 years post-collapse…Once you can smelt iron and copper, you can create wiring and start building electrical generators.
This fallacy is one of those ‘bad sci-fi’ tropes that comes up whenever this is discussed. Unless you’ve got a pretty out-there concept of what is going to cause ‘The Collapse’ (aliens suck out all the natural resources?) or forsee a Late Heavy Bombardment level event (which would mean no humans left anyway) we’ll be overwhelmed by the stuff we need to rebuild. Between the kitchenware and the pipes there’s probably 250lbs of cast iron in my house alone, millions if not billions of miles of copper wire still out there, and billions of motors of every variety from my shaver to the washing machine to the multi-ton ones in the dams up state. You won’t need factories (and required energy levels) that turn multiple tons of ore into pounds/kilos of usable metal, you’ll be turning pounds/kilos of already usable metal into something else usable. We’ll be just fine on the base metal front; the survivors are going to be wallowing in our leftovers for several centuries.
It’s like all those ‘apocalypse’ shows where 10 or 20 or 50 years after The Fall everyone is wearing buckskins and using bows and arrows. Silly. There are unfathomably vast amounts of guns, ammo and clothes sitting on shelves right now. And each one will last a long time with a little care; working 100 year old guns aren’t unusual and they’re using ammo more than 50 years old in Ukraine today. And any given Costco probably has enough clothes in stock for someone to change to new ones every day for a century, assuming they’re kept dry and away from vermin.
Getting from there to semiconductors and Isabel software is a lot harder.
We got from the industrial revolution to the nVidia H100 in, what, something a bit more than 150 years, and that included having to figure out virtually everything about how to build them. I suspect 100 years post-Doomsday is enough time to recreate computers, given we already know how to do it.
I don’t think software will be useful post-collapse, period. The real problems will be the age old pre-industrial ones like growing enough food, dealing with diseases without access to medicines, and defending against enemies.
Computers are too finicky. We won’t be able to keep them running long without a power grid or the ability to manufacture or repair batteries and solar panels. And the computers themselves aren’t repairable — for that you’d have to go back to 1970s tech like discrete components and small-scale ICs in DIP packages.
I do think these sorts of engineering projects are super interesting: safeguarding against civilization collapse is a huge challenge, and I can’t help but wonder what the best approach would be.
If you are serious about what the future of computing after a collapse would be, assuming you aren’t blinded or looking for water? Scavenging PCs, and hoping you’re lucky enough to find whatever with a programming environment. Anything else that i.e. involves flashing microcontrollers involves infrastructure you likely won’t be able to find, owned by soft targets that have likely died and their knowledge with it.
As always I’m skeptical that this is actually going to be of use in a post-collapse environment, even one where “computers are still available but we can’t build new ones”; I see this as more of a prepper fantasy, albeit one that’s more benign than stockpiling guns in your cabin.
That said, it’s a really interesting idea, literally compiling the C kernel as part of the boot process!
The post didn’t go into detail (that I could find) on how their mini-C differs from ANSI C; I found more info here. Most of the differences seem reasonable, except for the way “&&” and “||” aren’t short-circuiting — that would drive me nuts, and IMO the “real” behavior should be pretty easy to implement; hopefully they’ll add it.
Thought exercise: can you think of any hypothetical software project that would be useful, and that would be foreseeable pre-collapse?
I suspect the answer is there are few to none, because predicting the future to the required degree is just too hard, way harder than writing software (the “fantasy” part of “prepper fantasy” is that the useful skills will be the ones you already have). Maybe the closest you can get is something like the Arctic Code Vault, which might be useful in the event that civilization re-establishes itself.
I do think these sorts of engineering projects are super interesting: safeguarding against civilization collapse is a huge challenge, and I can’t help but wonder what the best approach would be. I just hope the author (or anyone else) doesn’t get sucked in too deep, to the extent that their current quality-of-life suffers.
Most if not all software “post-collapse” is going to be fundamentally useless, because post-collapse I’m concerned about food, water, shelter and coordination. A word processor or an operating system is not going to help me with any of those.
It is nice to re-examine the basis of our computing stacks though. Wish it wasn’t filed so far under the “prepper fantasy” mindset though. It’s a neat thought experiment/premise, much like “if all you had was a machine monitor, how would you build up to something you could actually use.”
It might be more interesting to think about 100 years post-collapse. Getting back to early Victorian technology with knowledge from now but no infrastructure seems feasible. Even without coal, you can create a similar energy density (though not the same volume) with charcoal, though it’s very labour intensive. Once you can smelt iron and copper, you can create wiring and start building electrical generators. Building steam engines is actually harder because the tolerances need to be much higher: you can build a water wheel out of wood and if you can magnetise a couple of lumps of iron then you can generate power (though for a lot of uses you will want a steady voltage and that is a bit more complex, but can be achieved with some damming to regulate the water flow rate).
Getting from there to semiconductors and Isabel software is a lot harder. If you assume some computers survived and you can generate a supply that can power them, what can you do in advance that will minimise the time taken to building a usable fab?
I suspect that the lack of coal is a bigger problem than anything else for building a society that could do this. The Industrial Revolution was possible only because the number of people required to produce food dropped to a small percentage of the population. That, in turn, was possible only because they were replaced with engines powered by cheap energy (steam from coal). It’s possible that wind and water could replace that for small communities but it’s not clear that they’d be large enough to train the specialists required for microelectronics or modern medicine. Stockpiling spare parts for your nearest nuclear power station might be the best bet: if you can run it at 10% peak load, it might run long enough to power your new civilisation to the point that it can build other energy sources.
This fallacy is one of those ‘bad sci-fi’ tropes that comes up whenever this is discussed. Unless you’ve got a pretty out-there concept of what is going to cause ‘The Collapse’ (aliens suck out all the natural resources?) or forsee a Late Heavy Bombardment level event (which would mean no humans left anyway) we’ll be overwhelmed by the stuff we need to rebuild. Between the kitchenware and the pipes there’s probably 250lbs of cast iron in my house alone, millions if not billions of miles of copper wire still out there, and billions of motors of every variety from my shaver to the washing machine to the multi-ton ones in the dams up state. You won’t need factories (and required energy levels) that turn multiple tons of ore into pounds/kilos of usable metal, you’ll be turning pounds/kilos of already usable metal into something else usable. We’ll be just fine on the base metal front; the survivors are going to be wallowing in our leftovers for several centuries.
It’s like all those ‘apocalypse’ shows where 10 or 20 or 50 years after The Fall everyone is wearing buckskins and using bows and arrows. Silly. There are unfathomably vast amounts of guns, ammo and clothes sitting on shelves right now. And each one will last a long time with a little care; working 100 year old guns aren’t unusual and they’re using ammo more than 50 years old in Ukraine today. And any given Costco probably has enough clothes in stock for someone to change to new ones every day for a century, assuming they’re kept dry and away from vermin.
We got from the industrial revolution to the nVidia H100 in, what, something a bit more than 150 years, and that included having to figure out virtually everything about how to build them. I suspect 100 years post-Doomsday is enough time to recreate computers, given we already know how to do it.
I don’t think software will be useful post-collapse, period. The real problems will be the age old pre-industrial ones like growing enough food, dealing with diseases without access to medicines, and defending against enemies.
Computers are too finicky. We won’t be able to keep them running long without a power grid or the ability to manufacture or repair batteries and solar panels. And the computers themselves aren’t repairable — for that you’d have to go back to 1970s tech like discrete components and small-scale ICs in DIP packages.
If you are serious about what the future of computing after a collapse would be, assuming you aren’t blinded or looking for water? Scavenging PCs, and hoping you’re lucky enough to find whatever with a programming environment. Anything else that i.e. involves flashing microcontrollers involves infrastructure you likely won’t be able to find, owned by soft targets that have likely died and their knowledge with it.