Hacker News new | past | comments | ask | show | jobs | submit login

So, let me see if I understand this correctly: this is supposed to run on z80-based computers after an armageddon comes and all the other "modern" computers are out of business, so people start building them by scavenging parts. Ok.

So, first of all, how are you supposed to download this thing onto your homebrew computer, given that internet will most likely be down?

"But if the collapse magnitude is right, then this project will change the course of our history, which makes it worth trying."

Mmmh, I think the author is a bit on the hyperbolic side here. I'm quite sure that anyone that can design & assemble a z80 computer can quite comfortably code some basic utilities by himself just fine. All the others won't care a bit about your OS. Sorry if I sounded harsh, but I actually was.

You're supposed to download it now and keep around. Just like preppers do with MREs and ammunition.

I said this further up the thread, but I'd rather have a raspberry pi, some screen cribbed from a smartphone, and some flash sticks.

Why plan for less than the raspberry pi level?

Because the pi might be needed for more demanding workloads (networks, routing, sdr radio system). With this solution you can build low-tech control systems and programmable switches with a bit of logic where a raspi would be overkill.

But the doomsday scenarios aside, this is super useful as an educational device. It can teach people what computers actually are and how they operate on the lowest possible level.

This is what excites me about this project. Another commenter posted the NandToTetris course, which is along the same lines. Computers/software are so complicated now, so much understanding gets lost in the upper levels of abstraction. Everything just seems like black box magic.

Or you know, a laptop. I've got like 6 of them piling up dust and all they need is 19V. The oldest from 2008, the batteries on 3 have sadly failed being unused, everything else still works.

I've also got 2 desktops from 2003ish (Athlon64 and Pentium III), they probably work, too, although they're stored in a garage (along with some other stuff like VCRs and CRT displays).

Not to mention routers old and new, all running Linux.

Yeah, why plan for less? All I need to do is scavenge around my property :D

Though a garden, livestock and a greenhouse will be a much higher priority. No one needs any sort of computing when they can't eat.

> Or you know, a laptop. I've got like 6 of them piling up dust and all they need is 19V

Mechanical parts fail first. The keyboards on those laptops will be gone after a few years of use. The USB ports won't last much longer. How long will the thermal paste and internal fans last?

The point of the z80 is that they're cheap and relatively easy to build from scavenged parts.

Not sure where you scavenge parts for a z80, I think it's much easier to scavenge parts for PCs. Plus, laptops are way more fixable than you seem to think, with simple soldering. Fans can be kept going for years with grease. External keyboards are ubiquitous. It's strange how people here bang on about RPi when a simple laptop is even more flexible.

And I'm also thinking more of a "repository of knowledge" use, not just simple controllers.

> Not sure where you scavenge parts for a z80,

TI calculators, microcontrollers, and more:


> I think it's much easier to scavenge parts for PCs. Plus, laptops are way more fixable than you seem to think, with simple soldering. Fans can be kept going for years with grease.

Collapse OS is talking about timelines of a century or more. Non-mechanical computers are the only ones that will last that long without the supporting infrastructure.

It's just a few thousand Z80 instructions in assembly notation. You can print it on paper, store the listings somewhere safe and type the code in by hand just like in the old days.

Why is that possible in this fictional reality but stockpiling an RPi isn't?

Because it doesn't fit the given narrative: it's all about scavenging/recycling.

Your reality does make better sense, but doesn't make a good story.

There are vastly more Z80's in circulation that Raspberry Pis.

Also, in a worse-case scenario, it's much easier to build a Z80 by hand from individual transistors and/or logic gates¹ (which would still be absurdly difficult, but not impossible) than any ARM CPU (including the ones in any generation of RPi). In a slightly-less-than-worse-case scenario, it's much easier to build a Z80-based computer by hand (i.e. with a Z80 you pulled off some other piece of hardware) than your average ARM-based computer (including any RPi).

¹ "logic gates" not necessarily being transistor-based, either; one could take a cue from the guy building a 32-bit RISC-V machine with vacuum tubes: https://www.ludd.ltu.se/~ragge/vtc/

A Z80 has over 8500 transistors, so even though it's much simpler than a Raspberr Pi, I don't think you're going to solder together an 8000 transistor processor in a post apocalyptic world. Even the old 4004 has around 2300.

I'd guess that those 8500 transistors would be better used to build thousands of much simpler logic controllers to help automate infrastructure that's lost its computer control systems.


There are simpler "transistor computers" that might be more feasible to build from discrete components:


There are examples elsewhere in these comments (linked by myself and others) of CPUs with similar (if not greater) transistor counts soldered together by hand out of TTL chips or even out of individual discrete transistors.

There are multiple ways to skin a cat, though (perhaps literally; this would be the post-apocalypse, after all!), and you're right that there are numerous ways to put transistors to use besides building full-blown CPUs. One of the key advantages of a general-purpose CPU is that it's general purpose and can be made to do all sorts of different things, but there are certainly plenty of cases where that ain't necessary and you'd need a fraction of that capability at most.

Still, they'll probably go hand-in-hand. "Chips" are just discrete components, whether crafted from a single chunk of silicon or itself built out of discrete components and treated as a single discrete unit. Building a whole general-purpose CPU from individual transistors is much easier when those transistors are already arranged on a little board you can plug into your bigger board. Chances are that no matter if someone's building a whole CPU or something more special-purpose and limited, that someone will be doing so in terms of already-assembled-and-composable gates rather than transistors directly, if only for the sake of one's own sanity.


[1] https://en.wikipedia.org/wiki/ARM_architecture#Acorn_RISC_Ma...

says about 30.000 gates, which is 10.000 less than Motorola 68000 and arguably faster.

[2] https://en.wikipedia.org/wiki/Zarch

at least made us Amiga and AtariST geeks envious.

Furthermore there is 'Microsequencer' like described there:

[3] https://en.wikipedia.org/wiki/Microsequencer

and following up from there to for example

[4] http://www.microcorelabs.com/mcl65.html

which is a 6502 softcore with microsequencing applied

Alas... There are many options to choose from according to the available technology, tools & knowledge. One does not have to make an exact copy of something which made sense for arbitrary reasons, which don't necessarily apply when doing it from scratch under different circumstances.

> says about 30.000 gates, which is 10.000 less than Motorola 68000 and arguably faster.

The Z80's at 9,000 transistors (not sure how many gates, but almost certainly a fraction of that transistor count), so even the Acorn would be heavy in comparison. Still doable, though; just takes more time.

In terms of speed, it has less to do with transistor count and more about how close together you can get the transistors. Big, hand-wired CPUs tend to be slower than small single-chip ones just from the sheer latency differences between components.

> One does not have to make an exact copy of something which made sense for arbitrary reasons, which don't necessarily apply when doing it from scratch under different circumstances.

True, and I ain't saying one does. If we're at the point where we have to hand-wire replacements, though, it helps to have at least some degree of compatibility with the thing we're replacing. There are at least some schematics out there for building 8-bit CPUs from TTL chips¹², and I'd imagine those would all be viable candidates if we have to re-bootstrap our computational power and run out of other CPUs to tide us over in the meantime.

Ideally we should be working on CollapseOS equivalents/ports for as many CPUs as possible, so that we know that no matter what we're stuck with, there's always a way to repurpose it. Just as importantly, though, we should be hoarding copies of pinout/wiring diagrams, hardware manuals, etc. to make sure we have the knowhow on the hardware side, too.

¹ http://cpuville.com/Kits/8-bit-processor-kit.html - happens to be bus-compatible with the Z80, though not ISA-compatible as far as I can tell.

² http://mycpu.thtec.org/www-mycpu-eu/index1.htm - more "modern" features like Ethernet and VGA out, so a more likely candidate for general purpose computing if we really do run out of Z80s to scavenge

The flash in the Pi will eventually die. Other high density ICs may also die from the laws of thermodynamics.

I don’t think it’s any coincidence that this collapse brings computing down to an era which is many people’s personal computing heyday, and not a decade before or after.

I’d hazard a guess that 8-bit machines played a part in the author’s young life - first computer, first job, happiest childhood summer, last computer they felt in control of before they got annoyingly complex - something like that. And therefore a collapse ending right when the author would have useful skills but things wouldn’t be too hard, is the most fun one to imagine.

Computing was around 40 years before the 1980s and electricity for a hundred years, but who wants to try and rebuild room sized punched card machines for ballistic trajectory calculations, get greasy fingers on mechanical parts, or deal with HT electrical power supplies safely, yawn, no fun there. rPi the same - by then everyone can do it and author isn’t special, so whatever. It’s not different enough from right now.

There are a very large number of Z80s around today and in-use. They are dirt-cheap and just right for many embedded applications. More importantly, they're simple enough to grok and to wire by hand; so yeah, were I stripping components for a make-shift computer in a post-apocalyptic world, I might just go for a Z80.

He specifically speaks of a "distribution" phase as imminent collapse seems apparent, before "blackout". So everyone who knows about it would hypothetically get a copy, and perhaps begin stockpiling some relevant hardware.

(SD cards seem like a good commodity to stockpile here, as he supports them, but they're likely incredibly hard to manufacture post-collapse.)

It may take a while but you could go Altair-style, by making a computer that is programmed through physical switches and then go and input every byte one-by-one. Though some automation would help here (AFAIK some people used a dotted paper tape reader and only entered the tape loader code by hand and the rest of their programs were loaded via the tape).

It's just a high effort apocalypse larp.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact