Hey, I'm the one who made this! My video on it probably does a better job of explaining it than the github so I'd recommend checking that out: https://www.youtube.com/watch?v=y7VtSK23_Jg
If anyone has any questions about the engineering process/game itself I'd be happy to answer.
Does the system of electromagnets do anything to enforce legal vs illegal moves? Seems like another way for your untrustworthy friends to cheat. When your opponent's distracted, just put your queen anywhere you like.
Also, is there a rule about moving one piece at a time? I was surprised I didn't see anyone two-handing it. If it were me, my opening move would be to grab 4 pawns at one (2 in each hand) and shifting them all forward two spaces. That would have been my first instinct.
No. The intention with the electromagnets wasn't really to prevent malicious cheating, just mostly to enforce cooldowns since it's very easy to forget/miss the different color of lights. The games are generally very chaotic and due to user error there are usually a few incorrect touches throughout the game that have to be quickly corrected manually anyways, so enforcing legal moves wouldn't make the experience better.
As for moving multiple pieces at once, the rule I have is only one hand and one piece at a time. This goes a long way towards preventing accidental bumps into other pieces and weird board states where there are multiple pieces in the process of being moved at once.
I'd agree that the video is much more captivating. It's really an incredible and innovative piece of hardware. I am curious, did you ever arrange a meetup with Magnus?
I'd really love to see some high level players try it. Did you get anywhere with letting Hikaru have a go? I know that he said he would be interested when he watched your video.
I love it! I'm the one who recreated Kung Fu Chess at kfchess.com. Very impressed at you bringing it to life -- definitely at least 100x harder to do :)
How do you handle the interaction between 2 pieces being moved at the same time ? Can I dodge a capture by picking up a piece ? Say 2 rooks are facing each other, how do you handle a mutual capture attempt ?
I have solved this problem, but still in a turn based setting :
* Both player choose their move, and moves are resolved simultaneously
* A piece cannot be moved 2 turns in a row (discrete cooldown time)
* When 2 pieces land at the same place, they are both considered captured.
* If they move in straight lines, in opposite direction, they are also both captured.
* NO CAPTURE for pieces crossing each other path / knights can exchange their positions.
* No pat. A player can skip his turn.
* the goal is to capture the oposite king obviosuly, not to checkmake it.
With a low blitz time, it solves the same fundamental problem as RT chess, but it can only be played online.
From a few test game, it looks like basic chess structure is preserved. But then there is the question of finding the Nash equilibirum of the game ! 2by2-simultaneous-move-chess with both rook and king leads to some sort of rock-paper-scissor.
The transcript of the video doesn't really work that well as the intro text in GitHub. The tone of it feels haughty, before even it has entered into the main details of the project. My suggestion is to rewrite it toning down the things that work on video but not as standalone text.
For what it’s worth, I didn’t find it “haughty” or tonally inappropriate at all (I thought it was fun and cute, and the right kind of unserious tone for a hacking-for-fun project).
If we’re doing a line-level read, though, it did take me a lot of reading to understand what the project was. Nothing that couldn’t be addressed with one more massage of your subhead “a physical chess board without the concept of turns”!
Super cool project, incredible execution, and you’re so personable—thank you for this work and your video!
Thought about both of these, for various reasons they won't work. Trust me I spent a decent amount of time on this, I think objectively making the bottom transparent with thicker acrylic is the best solution, but motion tracking is the other viable option and I wanted to play with the cameras.
I was thinking about spending an hour or two hooking up a chatbot api and maybe even sending it live position/score data so it could customize it's trash talk, but in the end I just wanted to get the video out so didn't. Would have been funny though!
(I'm the person who made it) Yeah that's fair, but at least for me I wanted to focus on the interesting part of "making a foosball robot" which is the "foosball" part, not the "fiddle with a home built vision system that doesn't actually work" part. I realize this is a bit ironic given my channel name though haha (From Scratch).
Yeah, that's so the robot knows where the human is so it can accurately make shots. I didn't mention it in the video since it was more of an implementation detail and not too important to understand what's going on.
I just want to say that this is incredibly impressive on so many levels. Your technical skills are obviously amazing but I really love how you were able to put this all together into an entertaining and well-produced video that anyone could understand. If you ever want to get paid 1/4 of what I'm sure you'll make as an engineer, you'd make a great teacher ;)
Thanks, was definitely hard to strike a balance between being understandable for non-nerds but also convey all the technical difficulties/achievements, but I'm pretty happy with how it turned out .
That is amazing I understand how powerful of nerd skills it took to do all those steps. Maybe the most amazing one to me is that terraria mod to speed up their wiring code without changing any other behavior, and the combination of skills to do such nerdy stuff and also finish the project in a form that normies can appreciate and also making a popularization video of it. It's a very useful combo of skills and interests.
I guess I have a question, I know you did it yourself but it looks like you were in some community. Is it like a discord channel or was it some of your classmates or what?
It was pretty much just me by myself on this other than for the raycaster engine, which was done by a friend of mine near the end. I'm on a few Discord servers on Minecraft computing, but it's pretty vastly different and the biggest parallel is just with high level accelerator stuff.
> "ReLogic if you ever read this although I doubt it"
They should definitely pay you to add this. They can do it by just hiring you remote for a little while even if you are still a student and you add this wiring code. You speeded up their code so many orders of magnitude without even having access to it. Most people would have a hard time doing it even if they can see the source code.
Also I have a question, what do you think about RISC-V? I never cared about embedded code until recently when I saw that you can buy an ESP32 chip for like five dollars and it's as powerful as a computer from the 90s and it has bluetooth and wifi, I was like wtf. I saw the most recent version ESP32-S3 uses RISC-V, and the new Arduino R4 uses ESP32-S3. Also related to RISC-V I saw this Jim Keller youtube video where he is all-in on RISC-V https://www.youtube.com/watch?v=_MrGNlXRi9M
I think it has a lot of technical merit and I heavily appreciate the fact that it's open source which is why I used it here. That being said in terms of microcontrollers I don't think it's doing anything fundamentally that much better than ARM, most of it's competitive advantage just comes from the fact that it's open source.
Edit: just as a disclaimer I've never actually written any ARM assembly/silicon design so I'm probably not the most qualified to have an opinion on this
I'm the one who made this. It's rv32i, and yes the 2017 specs. It passes every rv32i test in the repository you sent, see the test/ folder for the riscof plugin I wrote. I also have CI in a docker to headlessly run Terraria and programmatically run all these tests whenever the world changes on the main branch.
I'm not throwing around fully compliant lightly here!
- What would you say is your biggest learning from this experiment?
- What were the Aha! moments you experienced while learning like this? For example, something you vaguely understood before but is now super clear to you.
I would say the RISC-V immediate encodings were like that, on first look they seemed completely random but as soon as I started implementing the decoder I realized they were geniusly crafted to make my life as easy as possible.
Not really, unless you're a complete dunce. Money can be obtained in various ways that don't require soul sucking work of letting your life slip away in computer chair staring at screens. Anyone with a decently average emotional IQ knows this.
If anyone has any questions about the engineering process/game itself I'd be happy to answer.