Welcome Back, Voyager

April 27, 2024 by Elliot Williams 59 Comments

In what is probably the longest-distance tech support operation in history, the Voyager mission team succeeded in hacking their way around some defective memory and convincing their space probe to send sensor data back to earth again. And for the record, Voyager is a 46-year old system at a distance of now 24 billion kilometers, 22.5 light-hours, from the earth.

While the time delay that distance implies must have made for quite a tense couple days of waiting between sending the patch and finding out if it worked, the age of the computers onboard probably actually helped, in a strange way. Because the code is old-school machine language, one absolutely has to know all the memory addresses where each subroutine starts and ends. You don’t call a function like do_something(); but rather by loading an address in memory and jumping to it.

This means that the ground crew, in principle, knows where every instruction lives. If they also knew where all of the busted memory cells were, it would be a “simple” programming exercise to jump around the bad bits, and re-write all of the subroutine calls accordingly if larger chunks had to be moved. By “simple”, I of course mean “incredibly high stakes, and you’d better make sure you’ve got it right the first time.”

In a way, it’s a fantastic testament to simpler systems that they were able to patch their code around the memory holes. Think about trying to do this with a modern operating system that uses address space layout randomization, for instance. Of course, the purpose there is to make hacking directly on the memory harder, and that’s the opposite of what you’d want in a space probe.

Nonetheless, it’s a testament to careful work and clever software hacking that they managed to get Voyager back online. May she send for another 46 years!

Why Learn Ancient Tech?

October 15, 2022 by Elliot Williams 68 Comments

The inner orbits of the Hackaday solar system have been vibrating with the announcement of the 2022 Hackaday Supercon badge. The short version of the story is that it’s a “retrocomputer”. But I think that’s somehow selling it short a little bit. The badge really is an introduction to machine language or maybe a programming puzzle, a ton of sweet blinky lights and clicky buttons, and what I think of as a full-stack hacking invitation.

Voja Antonic designed the virtual 4-bit machine that lives inside. What separates this machine from actual old computers is that everything that you might want to learn about its state is broken out to an LED on the front face, from the outputs of the low-level logic elements that compose the ALU to the RAM, to the decoder LEDs that do double-duty as a disassembler. You can see it all, and this makes it an unparalleled learning aid. Or at least it gives you a fighting chance.

So why would you want to learn a made-up machine language from a non-existent CPU? Tom Nardi and I were talking about our experiences on the podcast, and we both agreed that there’s something inexplicably magical about flipping bits, calling the simplest of computer operations into action, and nonetheless making it do your bidding. Or rather, it’s anti-magical, because what’s happening is the stripping away of metaphors and abstractions. Peering not just behind, but right through the curtain. You’re seeing what’s actually happening for once, from the bottom to the top.

As Voja wrote on the silkscreen on the back of the badge itself: “A programmer who has never coded 1s and 0s in machine language is like a child who has never run barefoot on the grass.” It’s not necessary, or maybe even relevant, but learning a complex machine in its entirety is simultaneously grounding and mind-expanding. It is simply an experience that you should have.

Another Kind Of “Bare Metal”: 6502 Computer Powers RPN Calculator

November 17, 2020 by Dan Maloney 11 Comments

[Mitsuru Yamada] states that one of the goals for this 6502 computer build was to make it strong enough to survive real-world usage. In that regard alone we’d call this a success; the die-cast aluminum enclosures used are a little blast from the past and lend a nice retro industrial look to the project. The main chassis of the computer fairly bristles with LEDs and chunky toggle switches for setting the data and address busses. The interior is no less tidy, with the 6502 microprocessor — date code from 1995 — and associated support chips neatly arranged on perf board. The construction method is wire wrapping, in keeping with the old-school look and feel. Even the hand-drawn schematic is a work of art — shades of [Forrest Mims].

As for programming, this machine is as low-level as it gets. Nothing but 6502 machine language here, entered manually with the toggle switches, or via an externally programmed ROM. The machine can only address 1k of memory, a limit which the code to support the RPN calculator add-on [Yamada] also built brushes up against, at 992 bytes. The calculator keypad has a 20-key matrix pad and an eight-digit dot-matrix LED display, and can do the four basic operations on fixed-point binary-coded decimal inputs. The brief video below shows the calculator in action.

We love the look of this build and we’re eager to see more like it. We’ve seen a ton of 6502 builds from discrete chips lately, and while we love those too, it’s nice to see one of the big old DIPs put back in action for a change.

Continue reading “Another Kind Of “Bare Metal”: 6502 Computer Powers RPN Calculator” →

Riding The Nostalgia Train With A 6502 From The Ground Up

September 19, 2019 by Dan Maloney 44 Comments

In the very early days of the PC revolution the only way to have a computer was to build one, sometimes from a kit but often from scratch. For the young, impoverished hobbyist, leafing through the pages of Popular Electronics was difficult, knowing that the revolution was passing you by. And just like that, the days of homebrewing drew to a close, forced into irrelevance by commodity beige boxes. Computing for normies had arrived.

Many of the homebrewers-that-never-were are now looking back at this time with the powerful combination of nostalgia and disposable income, and projects such as [Ben Eater]’s scratch-built 6502 computer are set to scratch the old itch. The video below introduces not only the how-to part of building a computer from scratch, but the whys and wherefores as well. Instead of just showing us how to wire up a microprocessor and its supporting chips, [Ben] starts with the two most basic things: a 6502 and its datasheet. He shows what pins do what, which ones to make high, and which ones get forced low. Clocked with a custom 555 circuit that lets him single-step and monitored with an Arduino Mega-based logic analyzer, we get a complete look at the fetch and execute cycle of a simple, hard-wired program at the pin level.

This is one of those rare videos that was over too soon and left us looking for more. [Ben] promises a follow-up to add a ROM chip and a more complex program, and we can’t wait to see that. He’s selling kits so you can build along if you don’t already have the parts. There seems to be a lot of interest in 6502 builds lately, some more practical than others. Seems like a good time to hop on the bandwagon.

Continue reading “Riding The Nostalgia Train With A 6502 From The Ground Up” →

Getting To Know Every Bit Of An ATtiny13

June 2, 2019 by Elliot Williams 15 Comments

We recently heard it said of a hacker who pulled off a particularly nice VGA hack on an 8-bit microcontroller: “He knows all the bits, personally.” High praise, indeed. If you want to get on a first-name basis with a ton of transistors, then have a look at [Heinz D]’s Vacation Course in ATtiny13 Assembler (original in German, translated into English by robots here).

But be warned, this isn’t the easy way to learn AVRs. Not content with simply stripping away every layer of abstraction, this month-long “course” in AVR assembly starts off programming the chip initially with just two pushbuttons in its native machine language of high and low voltages. But still, especially if you can get a few assignments done in one sitting, you’re writing in the relative splendor of assembly language and uploading code with a proper programmer before long, because there’s a real limit to how much code one can toggle in before going mad.

There’s a beautiful minimalism to this entirely ground-up approach, and maybe it’s an appropriate starting point for learning how the machine works at its lowest level. At any rate, you’ll be able to lord it over the Arduino crew that you were able to get blink.ino up and running with just a pair of mechanical contacts and a battery. Real programmers…

And once you’ve mastered AVR assembly language, you can recycle those two buttons to learn I2C or SPI. What other protocols are there that don’t have prohibitive timeouts? What’s the craziest code that you’ve ever entered bit by bit?