Hacker News Reader

Comment by nwah1 | Hacker News Reader

Why does it need to contain an actual ARM processor?

Need to? It doesn't. Why might users care? Well, the kinds of people who are into custom keyboards often use custom firmware too... while this keyboard advertises QMK support, if I don't like QMK and want to use something else, which chips are inside matters.

Making this even more concrete, I am in the market for some new keyboards, and at least one of them is one I'd like to write my own firmware for. Rust is far more supported on ARM chips than on some of the other popular chips that are used to build keyboards, and so knowing this has one makes me more likely to buy it than others. I don't want this specific keyboard, so it isn't a win for me, but there are surely dozens of folks like me, dozens!

I think OP meant "Why does this keyboard need so much computing power". Most keyboards you can program use an AVR chip, which is more than enough, and you can also program that from Rust.

The jump from that to this ARMv5 CPU is similar to that from this ARMv5 to an Intel Xeon. And sure enough, if someone were to build a keyboard with an Intel Xeon, you could also program that from Rust. But the question still remains: why would anybody do that ?

EDIT: oops, you edited substantially from when I replied with what's below.

------------

That's correct, yes. I guess I didn't say that explicitly, but that's another decent reason to advertise this, it's not exactly the norm.

(And for my personal quest, the latest Rust nightlies do work on AVR!)

So, this is 100% speculation, but....

Specifically, this has a STM32F303xC. Looks like (depending on quantity, of course), you're looking at $3 to $5. Their other well-known keyboard, the Ergodox EZ, uses an ATMEGA32U4-AU. It's also $3 or $4 on Mouser. I don't know what the actual BOM looks like in total, but it seems like this isn't significantly different in price? Especially at the likely volumes here?

I'm still new to this game, so maybe I'm missing something.

Thanks, this is an interesting discussion. I thought the arms would be much more expensive: https://www.reddit.com/r/olkb/comments/6wt6mi/evaluating_mcu...

Nice, thanks for that link!

Actually, there are people who have already written firmware in Rust for their DIY/kit keyboards. Nothing yet mainstream like QMK. Here's a write up of one of them: https://josh.robsonchase.com/rest-of-the-keyboard/

Meet the Ferris keeb that was just recently developed. Currently built using QMK but will eventually have Rust firmware. https://github.com/pierrechevalier83/ferris

I know some people have, but part of the fun is doing it yourself :)

There's also https://github.com/TeXitoi/keyberon

Indeed!

Was just looking at that yesterday.

While Cortex-Ms are surprisingly powerful compared to the microcontrollers of yesteryear, they're still targeting the same segment. They are not comparable (in terms of cost, power consumption, performance...) to the ARM in your smartphone, let alone the ARM in AWS Graviton. They have kilobytes, not gigabytes of RAM. They run at double digit megahertz clocks.

I would hazard a guess because it allows a ton of configuration and key remapping options.

It could use an AVR but you don’t save much and the world is on ARM. You would “need” an ATMega 32u4 anyways for hardware USB support. A purposed KB IC probably wouldn’t handle split halves.

Why not? They are affordable, accessible and common microcontrollers. Where you might have used an AVR or PIC before, nowadays a small ARM is also a completely normal choice.

Reply on news.ycombinator.com