Progress since the 1980s and 1990s

Did anyone else here grow up in the previous century? So many things have changed and improved since then, I don’t even want to start listing them all. In the area of keyboard ergonomics we also have a lot of progress: there are a lot of options of split keyboards, many of them also with finger-friendly column-staggering. But what really bugs me is that those are mostly niche products. Unlike smartphones, TV streaming, AI chat, which have all achieved mainstream usage. Today’s “standard keyboard” is still almost the same as a 1990s standard keyboard.

And I am talking especially about the layout of the keyboard, something which doesn’t need any new technology. It just needs a change of culture to value usability more. It’s the same change of culture that has happened in software companies and IT departments: in the 00s, UX was not a well-known term and design a minor concern; however in recent years, every software team for consumer applications had to have a UX expert. And consequently, software became easier to use. Why did the same not happen for keyboards?

I couldn’t let go of this question, so I looked around for different keyboard designs and the history of the ANSI/ISO standard… and I found that before the standard was set, there were actually a lot of different designs on the market and some of them actually had some better design elements. So I took some of those design elements and recombined them into a 21st century keyboard layout that removes some of the worst problems of the ANSI/ISO standard keyboards: too much load on the right pinky, not using the thumb’s potential, and wasting space with oversized keys. Basically, it all went into the direction of the Planck/Preonic keyboards that have perfect symmetry and only 1u keys… gaining some of their advantages, but still keeping the most familiar features of the traditional layouts.

I call it the Thumbs Up 16/5 keyboard layout:

As you can see from the key size chart below, it is 16 key units wide, just like an ANSI/ISO 65% or 75% layout. It’s the same form-factor that a lot of people already have on their desk, but providing more keys and making them more accessible in the center of the keyboard. While it does require a little effort for getting used to, I think this is about the same effort that one has to invest for getting used to a new laptop keyboard – because those have their navigation and other keys also in a variety of positions. The Thumbs Up, on the other hand, is designed intentionally to work both as a laptop keyboard layout and as a stand-alone model.

What strikes me is that I had to do nothing crazy new to create this layout:

  • the key sizes on the left edge (Backspace, Return, and the key between both) are taken from Apple’s keyboard layout in use since the actual 1980s.
  • the key sizes on the right are just doing the same modification: shaving 0.5u from all keys, so that the stagger is unchanged and the keyboard whole number of keys again (Apple’s layout is only 14.5u wide).
  • finally, the right half of letters and numbers move by one key to the right, something that Colemak-keymappers have done in software for years; now we make it official.
  • the above changes together create two “new” columns of keys in the center: one with the moved character keys from the right, and the other created from down-sizing keys, so it can be used for extra keys that are missing on standard 60% boards, such as `~, Delete/Insert, and navigation keys.
  • and since we have that middle part of the keyboard now, we can as well add another column in the center, widening the whole keyboard from 15u to 16u and place the arrow cluster in the middle as well. This creates some nice gaps to find the arrows by feel while still keeping all keys in the rectangular box.

While there’s no fixed standard for n-percent keyboard key counts, the Thumbs Up 16/5 has exactly 72 keys. I consider the key labels shown above only an inspirational example. If you don’t need an Insert or Delete or Home/End keys, use them for macros or shortcuts. I think it’s great to have a default keymap that makes sense and optionally a lot of keys up for customization. (That’s also why I created the example picture above with flat keycaps: I want to make it easy to switch to other letter maps, like Colemak, or even switch around some of the other keys just as it fits each user best. The layout should be a pattern of sensible key sizes, basically the shape of the switch plate, and mapping the keys left to the user or seller of the board).

Anyway, I didn’t write this post to advertise my own ideas. All I wanted to show is that there’s a lot of room for improvement on the old typewriter-based keyboard layouts. I wish I could buy such a keyboard, even if I had to assemble it from parts and even if I had to buy lots of keycap sets to extract enough of each type of >1u key. What I don’t want to make is a PCB and switch plate from scratch.

Does anyone else miss layout innovation in the mass market? (Apart from different ways to split Shift keys…)

3 Likes

If it works for you, then it works, full stop. There’s also some good stuff here for others. Fundamentally, you’ve got a unibody split layout here, but you’ve (prudently, IMHO) filled in the central area with stuff that’s useful to people who are not obsessed with keeping their hands stuck on the home row. I actually think that’s kind of good for ergonomics. So many of the designs do a great job on the “S” part of RSI, but moving around a bit can help with the “R.” Gotta say, I’m also liking the idea of no stabs. :slight_smile:

If you can come up with a way to shave it down to 15u without compromising your goals, that would open a whole world of existing 60% cases and make a really professional-looking build very doable, especially with Yuzu/FKcaps. I have made some boards with layouts that use a few similar ideas (though the general philosophy is a little different), and I have found, poor typist that I remain, that saving 1u or so by eliminating the stagger on the numrow has been worthwhile. You might also take a peek at “katana” layouts and see if that’s something that strikes you as useful.

Frankly, I have a few nice aluminum rectangles now, and I’m quite fond of them, but it’s the deep customization and DIY spirit that sucked me down this rabbit hole and I love to see what people come up with. If you haven’t already, I’m sure you’ll soon be frequenting kbd.news – so many interesting builds there.

Edit: and a peek at your github shows you obviously have done plenty of research, LOL.

2 Likes

Uh, thanks for the kind words :heart: .

I actually have a 15u version of this, it was even the first one I made:

But then I found that outside the enthusiast circles, there are much more keyboards with arrow clusters on the market and they’re all 65%. I even looked for some Keychron low-profile boards to buy and replace their innards (keeping case, switches, most of the keycaps), and found that the 60% variants were usually out of stock. What makes it even more complicated is that I really like low-profile keyboards (like my Iris CE and the typical laptop keyboard height), and that market isn’t served by custom cases. That’s what gave me the idea of replacing plate and PCB in an off-the-shelf keyboard… but still I’d need somebody to design and produce those two parts. My apartment is already a bicycle workshop and it should not also become a keyboard parts factory.

During my research I have become kind-of an expert on keyboard layouts, but I don’t also want to become an expert in keyboard parts. I have a strong feel that more people would like to have a keyboard with such a layout, but when it comes to real keyboards there are so many factors at play: low vs high profile, mount type, etc. Interested people would have to decide on a first set of specifications for a PCB to be made. My current idea is a tray-mount PCB, but there still would need to be separate low-profile and standard MX variants. And if designing for existing cases for both variants, then the case-interface (screw-holes, USB-port) would also be different for both. And after deciding for low-profile, I guess we’d have to settle for a specific case, because there’s not even a semi-standard like the Tofu65…


PS: yes, the Katana was an inspiration! I might as well post the link to my little research app: Keyboard Layout Visualizer - Compare QWERTY, Colemak, and others on very different kinds of keyboards, including the highly ergonomic split orthogonal. It started, because I wanted to decide which letter mapping to use on my new split ergo board… but as soon as I had decided that question, it became more about the actual physical layouts. Since there are a lot of other web sites about letter maps already, doing something with layouts is actually more interesting…

1 Like

If you haven’t yet, I would strongly recommend you seek out the 40s Discord server. A lot of the DIY spirit of the community has migrated to the 40s space, and that includes people with small-scale fabrication experience and connections. Seems like you’d fit right in. As for a 15u layout, I guess it all depends on how convincned you are that you have your bottom row where it needs to be. Something like this might be helpful, and with an arrow cluster you might find it livable to redirect Home/End and PgUp/PgDn to a layer, like Apple laptops do. Just a thought. You’ve clearly put a lot of research in and have good reasons for the choices you’ve made.