Comment by jrockway | Hacker News Reader

Comment by jrockway | Hacker News Reader

Their website won't load, but it looks like they've been working on a prototype of a keyboard that has e-ink displays on each switch since around 2016.

This has been attempted before: https://en.wikipedia.org/wiki/Optimus_Maximus_keyboard but with LCDs (or maybe they were OLEDs, since all the reviewers complain about burn-in). The reviews were overwhelmingly negative and you can pick one up cheap used because nobody wants them.

Overall, I find these designs awfully expensive on many fronts. Looking at the keyboard is a huge productivity sink, and if you stop looking at the keyboard, you won't care what the legends look like. And, having 101 screens attached to your keyboard is pretty costly compared to having 0 screens. I think these make for a good conversation piece, but I'm not sure you'd prefer typing on them to a keyboard with a better layout and better switches. And if you want a macro pad, just get a stream deck, which you can buy with same-day shipping instead of crowdfunding it. (Or just build your own. There are tons of kits out there that cost on the order of $10. No per-key screens though.)

I could see them being useful for keyboard designs with very few keys, like https://olkb.com/collections/planck.

If you have deeply-nested layers, having some way to identify what state your keyboard is in can be helpful. Having screens instead of LEDs is definitely an expensive luxury, though.

True. Personally I have 3 LEDs on my keyboard that I use to show the current layer (in binary). I then carefully numbered my layers so that I use 0, 1, 2, and 4 in preference to others ;)

That’s an interesting idea, actually. A keyboard that could visually identify state might be useful for education: say a typing or Vim tutor. It’d be kind of nifty if a key started flashing so that you could identify its location (for learning a layout), or had a little icon suggesting “insert” or whatever command it triggers. T’would be an expensive teaching tool, though.

Art.lebedev had full-size keyboards, as well as smaller pads with 3, 6 or 15 keys.

My optimal keyboard would have: mechanical switches, per-key RGB LEDs, MX keycaps with translucent borders to see the LED lights ("pudding" style), per-key legends that would change with the keyboard language and layers.

I only want to be able to look at the keyboard when I'm not familiar with the layout, but I want to be able to do that.

I am just not convinced that an O(n) visual scan over the keyboard is that useful. What programs really need is a good "search" function for commands. Obviously scanning over an entire toolbar is useless, but if you can search named commands and the application shows you where in the menu the command is and what its shortcut is, then that would be better than a keyboard that just puts toolbar icons on the keys.

(I have used variants of this feature in many applications and love it. There are a bunch of functions in Fusion 360 whose name I know, but not where in the menus it is. But they have a search-and-invoke UI, so I can run them. I do wish it would show me where in the menus they are, though, because after the first couple times it sure is inefficient.)

> having 101 screens attached to your keyboard is pretty costly

I wonder if it would be somehow feasible to cut costs by using a projection scheme.

Perhaps each key could be essentially passive rear projector screen, then you'd have a system of mirrors, lenses, prisms, or whatever below the keys so that one DLP or other projection system could hit all the keys.

Maybe having 1 active component and 101 passive components (hopefully all plastic) would make the whole thing a lot cheaper.

Of course, there would be the small matter of building it so the geometry works out where you can have the image hit the key and also have a switch and spring beneath it.

I'm one of the founders, the first unit used a prototype EPD from E Ink Holdings, a custom display with around $500k in NRE was developed for a Macbook size keyboard module. This single large display with a magnetic mechanism operating around the footprint meant the total unit cost in COGS was significantly lower than $100USD at scale. The entire mechanism was only a few mm in height including the display. The travel and tactile profile was designed to a butterfly switch mechanism spec.

Your keyboard is neat, but I just don't think it's for me. If you've ever looked through menus for some command, you know that it never feels good. Looking down at your keyboard and doing the same visual search would feel just as bad.

If I were spending on keyboard research and development, I'd want a silent design that feels as tactile as buckling spring switches. That would benefit me (and anyone nearby) every time I press a key, which is thousands of time per day. A display on each key? Neat, but that's about it.

Reply on news.ycombinator.com