▲ 133 points
back
31 comments
I 3D printed some Zuse-style mechanical NAND gates last year, that was a lot of fun and makes for a great fidget toy:
https://mero.ng/i/vMdqQYJG.jpg
I've kind of given some thought of doing a higher-quality metal production run of these with nice finish and engravings of the inputs and outputs as a geeky desk fidget.
Model credit goes here (and of course to the original Zuse patent application the design is from), although I made a NAND remix as I wanted a universal gate and fixed up all of the tolerances and still really have to upload my version:
https://www.printables.com/model/69642-zuse-inspired-z1-logi...
So cool! Should link to the project homepage http://www.nablaman.com/relay/ Or the news page with pics & videos. http://www.nablaman.com/relay/progress.php Last update was in 2011, btw.
There was an update in November 2013 at the end of https://www.nablaman.com/relay/story.php
> The last time I wrote something here was two and a half years ago. Since then, Zusie has been more or less finished. Also, I've moved to a different city, and Zusie has followed along.
I think that's the fate of every project like that. Not to take away from it, it's still fantastic and geeky. But at some point, you have this realization that you're pouring years of your life into something that is not only utterly useless, but also won't be appreciated by others. You can't even play Pong on it. There's literally nothing you can demo to non-nerds, other than "this device makes looks complicated and makes noise".
Maybe you'll make it to HN more than a decade later, but that's it. No one will buy it from you. Your heirs will toss it out. It... sucks.
The real project is the sense of accomplishment we gained along the way, or something like that.
But anyway, there are institutions like the Craftsmanship Museum <https://craftsmanshipmuseum.com/> that exist to present this kind of passion project to the interested public. That one in particular came out of and is still very centered on the hobby machinist and model steam engine community... if there's not already something similar for electronics and computing type projects, it definitely seems like maybe there could/should be.
> The result is largely who you became while you build it, not the physical end product.
And yet you've developed this knowledge for yourself that applies to other things. Among them you develop probably the most imperative skill which practically nobody focuses on; problem solving. And if you've got kids that share your interests they get that knowledge and experience also.
Humans need hobbies. Humans need to explore their curiosities. If you're not doing this you're hiking or biking or gaming or going to the gym or kayaking or.... something equally useless.
I really feel this statement. I'm being forced out of a company that is struggling to pay my salary, after almost 18 years. This is a career where I taught myself and continued (and continue) to improve. I know that my best skill is the ability to solve unique problems, and that's mostly been from needing to complete a project or the business loses their time and money. For most of my career, I have been the sole developer, and impostor syndrome is real and haunting. Having HN, and following various news and blog sources, I think I am in a good place with my knowledge. It's just transferring to another company after almost two decades that is the scary part.
Having a senior position, but still very much wanting to be involved in coding and architecture, makes me nervous that I won't be able to keep my same lifestyle without going into full-time management. Problem solving is often tossed to the side for a "good enough" answer. I understand that as a business need, but when things start to get really complicated, pulling in an existing library or solution is often not possible.
Sure, but we're herd animals and also need validation, even if we're pretending not to. We go on a journey of self-discovery to Nepal so that we can talk to others how profound it was, etc.
I'm not dissing hobbies, I'm just saying that building stuff like that is a very lonely hobby, which is why such projects almost always fizzle out.
Who's to say no one appreciates it? And even if it were unappreciated, that doesn't make it any less worthwhile to the person working on it if that's what they choose to do with their time. Not every project needs to live forever with a bustling community and updates. You can quietly work on something, finish it one day, be proud of your work, and put it online for others to admire (or not). That's plenty enough of a purpose and reason to build these things IMO.
That's beautiful and mesmerizing - video here: http://www.nablaman.com/relay/progress.php
I built one of these[0] a few years ago, and really enjoyed the build process and playing around with it. It was a pretty big project, but no special skills required.
The creator of relaysbc posts here every so often.
The bare PCB is here https://www.tindie.com/products/jhallen/single-board-relay-c...
I have not completed mine yet, what was the hardest part during assembly or the gotcha you had to debug?
I think the hardest part was trying to clean the flux off of the board, and I never did get it that clean.
I chose to use sockets for the relays, and at first it had some weirdly inconsistent behavior. Removing and re-installing all the relays fixed it, though.
I went very slowly and checked my work as I went. I don't understand the circuit well enough to debug it, so I was very motivated to work accurately.
Shameless plug for my relay computer: https://github.com/artemonster/relay-cpu
Related. Others?
Zusie – My Relay Computer - https://news.ycombinator.com/item?id=8711869 - Dec 2014 (21 comments)
As of late I've been watching a Youtube creator, DiPDoT[0], building a relay computer.
DiPTDoT's project is still in-progress. His computer has an 8-bit data bus and a 16-bit address bus[1], like the Zusie. His registers are a little more generous than the Zusie.
I really enjoy the sound of DiPDoT's test rigs running his cards thru functional testing. The rhythm of a relaying computer running code with loops is probably pretty fun.
One of the most polished diy relay computers (and video series) I've seen is by Paul Law [0]. He's been working on it for over 10 years, and has just started a series on using it to calculate Pi.
the guy uses same relays as I do and he solders them onto pcb. Ive learned my lesson the hard way that these relays are unreliable and they do die, so in my design all of relays are socketed. hope the guy will adjust his design too
Do you buy the sockets or do you have a custom solution for that?
these are just simple DIL sockets
This is a very cool project!
I wish this was around when I was teaching computer architecture. I use to challenge students to think outside the box when it came to switchs. One example is to build a computer out of paperclips.[1] The book can be found on eBay, Amazon and other used/old book outlets.
[1] "How to build a working digital computer", Edward Alcosser, 1968
Another relay computer for those interested, with hand built DRAM!
http://www.northdownfarm.co.uk/rory/tim/tim-8.htm
> 8-bit data bus and 16-bit address bus
What? That is very intense. If I were doing relays, I would have gone for 2 bits or something, 4 bits max. Hardcore.
Agreed, 8-bit data, 16-bit address (but 12-bit program counters) is massive for relays! That's a lot of relays to do anything. But it turns out their start was when they stumbled on 1500 relays in good condition. So they were set in relays from the start and could aim "rather" big. They do have a lot of relays.
Having seen this in person my favorite aspect is the sound it makes, absolutely mesmerizing.
I'd love to see, and hear, a video of the computer in operation
I see "This video displays Zusie the Relay Computer...", but there's no actual video to see.
https://youtu.be/-ReqdyCxZ9I?si=e6WjONWMOGHG7Wja
Sounds fantastic on YouTube, even better IRL.
There is now a whole bunch of relay computers on youtube! It's become easier to do it seems. This one 2011, some more recent on youtube.
(2011)?