Opinion on Small Keyboards and Big Spacebars

Posted on June 29, 2023 by Richard Goulter
Tags:

The mechanical keyboard hobby pretty much exploded in popularity during the pandemic.

“Mechanical keyboards” generally brings to mind the loud, clicky switches; or perhaps people who have spent a lot of time (and often money) in order to get a keyboard that feels nice.

But, there’s a niche part of the hobby which is passionate about playing around with more exotic keyboard designs.

Rather. The mechanical keyboard switches allow for manufacturing low volumes of keyboards which improve upon the standard keyboard design.

The standard keyboard design suffers several problems, which make the keyboard less than efficient:

  1. The keyboard’s spacebar is impractically big.

  2. The hands have to be close together, rather than at more natural distances apart.

  3. It’s asymmetrical.

  4. The rows are staggered (again, asymmetrically).

  5. There are keys of different sizes.

For myself, although I can understand spending more money to get a ‘nicer feeling keyboard’, I cannot understand why anyone would pay luxury prices for keyboards which still suffer these problems.

(“Cheap Logitech keyboard” gets to answer these by way of: this is what keyboards have always been and what people are used to. With economies of scale, you get to have a very cheap keyboard).

But to explain why I think each of the above is a design flaw, and how custom mechanical keyboards tend to improve on these:

Design Flaw: The Spacebar is Impractically Big

The spacebar is impractically big. It’s often 6 times as large as the normal keys.

Or to put it another way: our thumbs are much stronger than our pinkies. But on a standard keyboard, our thumbs pretty much only get to use the spacebar, and the pinkies tend to get used frequently for all sorts of important keys (such as backspace, shift, enter, etc.).

There’s no good reason for the spacebar to be that big: although not everyone presses the spacebar in the same place, everyone will tend to push the spacebar consistently in the same place they press it. (As opposed to using the full with of the spacebar).

It would be better to replace the wide spacebar with a key that’s as wide as the standard keys; this would allow putting another 5 keys in the same space. – It wouldn’t matter what those keys did; anything would be a big improvement.

Multiple Thumb Keys Hugely Increase What Your Hands Can Do Without Moving

The term “home row” refers to the row of keys where your hand rests when it’s not typing. The ‘F’ and ‘J’ keys have bumps on them which allow your hands to find their place, without you having to look for them.

If instead of one big spacebar, there were 6 individual keys, then your thumbs would be able to use 2-3 keys each.

e.g. I like using these for Space, Backspace, Enter, Esc, Tab, Del. (As a developer, I use the Esc key frequently).

Whereas, with the spacebar being so big, in order to hit those keys, you’d have to either stretch your pinky, or move your hand from home row.

Thumb Keys are Great Layer Keys

“Layers” allow for re-using a key for multiple purposes.

This idea is extremely common on laptop keyboards: there’s often a “Fn” key, and the keys will indicate what functionality they have if pressed when the Fn key is held. e.g. “Volume Up” / “Volume Down” might not have their own keys, but might be available when a Fn key is pressed.

That’s an example of using keymap layers.

From a pessimistic perspective, it’s a technique used by small keyboards to cope with not having so many keys.

From an optimistic perspective, it allows more functionality without having to have more keys.

Thumb keys are excellent keys to use for as “Fn” keys for layers.

Aside: Dual-Use Keys are Neat

One idea in the keyboard space which is useful for really getting the most out of a keyboard is “key behaves like T when tapped, like H when held down”.

It’s relatively common for programmers to come across this good idea: When using Vim, programmers want to tap Esc all the time; when using MacOS, programmers want to hold Ctrl all the time. Programmers pretty much don’t want to use CapsLock. (Not that frequently, anyway).
So, solution: use the CapsLock key as an Esc key when it’s tapped, but have it act as a Ctrl key when it’s held down.

Above, I was pointing out that the weak-pinky finger gets used for Shift, Ctrl, Win, Alt keys.. one neat idea to alleviate that is to put these modifier keys underneath the F,D,S,A keys (and J,K,L,; keys) using this dual-use tap-hold functionality. – This technique is known as “home row modifiers”, or “home row mods”; since it allows using the modifier keys without having to leave home row.

Similarly. Above I argue that thumb keys are well placed for backspace, and are well placed for use with layers. Indeed, I think having these as dual use is a good idea, too.

Design Flaw: The Hands are Too Close Together

While this is (to an extent) unavoidable with small devices like laptops, what I mean is that for desktops, keyboards can be in “split” halves.

This allows positioning each half of the keyboard so that the hands can rest at a more natural distance.

I’ve seen for many people, this is a significant improvement. I’m not as won over.

I also see there are many keyboards which are “split”, but retain all the other design flaws mentioned above.

Anyway. Split keyboard designs are quite popular amongst the niche of mechanical keyboard enthusiasts that prefer exotic designs.

Design Flaw: The Keyboard is Asymmetrical

The standard keyboard is asymmetrically laid out. With the hands resting on home row, there are far more keys to the right than to the left. That’s particularly strange, given that most people use a mouse to the right of the keyboard.

Even laptop keyboards suffer the same. – Worse, it’s easy to come across laptop keyboards which don’t have all four modifier keys (Shift, Ctrl, Win, Alt) on the right hand side.

At a glance, it might look like the key map is symmetrical; but, the asymmetry is clear when considering that “Q” is the left hand pinky’s, and has one key to its left; whereas “P” is the right hand pinky’s, and has 3-4 to its right.

To me, it doesn’t make sense why you’d introduce that asymmetry. – Why add keys to one side, but not to the other?

Design Flaw: The Key Rows Are Staggered

This somewhat goes hand in hand with the above point.

The Katana60 one of one a very few keyboards I’m aware of that are symmetrical while retaining the row stagger. (It aims to use keys from a standard keyset; so, I guess I’ll give it that).

Most of what I dislike about row stagger could be blamed on asymmetry: it hardly makes sense for the left hand’s keys to “slant left” with the same slant that the right hand has. – I mean, “R” relative to “F” is the same as “U” relative to “J”; but, the slants should be mirrored.

What does seem more sensible is either to lay the keys out in a grid, or to stagger the columns. (I’ve also seen keys “splayed” out).

The grid arrangement is called “ortholinear”. (Despite that “ortho” and “linear” both mean ‘straight’ in Greek, Latin).

Staggering the columns means that each column of keys is straight, but each column is staggered compared to the one next to it.

These all try and deal with how the fingers reach the keys on a flat keyboard. Perhaps something like a Maltron or a Dactyl Manuform are much better than either of these.
– For flat keyboards, I’m willing to compromise.

95% of what makes a row stagger bad is the asymmetry. But, my intuition is still that there are good reasons row stagger is much rarer once breaking from the traditional shape.

Design Flaw: The Keys are Different Sizes

This is arguably a consequence of the above design flaws.

But I’m also not sure why you’d design keyboards where the keys had different sizes?

Rather.

One consequence of keys having different sizes is that keycap sets may not be compatible with every keyboard.

This affects the exotic keyboards I like, which only have space for “1U” keys.

But I see that keycap sets will often have the same modifier keys in an additional two or three different sizes. – It seems wasteful and silly; or it would, if all keys were “1U” big all the time.

Aside: Comfort is the Main Goal, Not Speed

I think the most common reaction people have in discussions about alternative keyboard stuff like ergonomic layouts, or the Dvorak layout, is asking whether such alternatives are faster.

Perhaps the thought process is: “I’m fine with my keyboard as-is; but, sure, if I could type faster then it would be worth putting in effort to learn a new thing”.

The most common comment I do see from people who have used these alternative methods (different keyboard layout, etc.) is that they are much more comfortable.

And regarding the above discussion about improving on the design flaws of the standard keyboard, the improvements are all about being able to do more with your hands without having to leave home row.

Keyboard Recommendations

So what would I recommend to those who are looking for a keyboard which doesn’t suffer these flaws?

For desktop use, a good default recommendation is ZSA’s Moonlander.

If you’re rarely using mouse + keyboard work (e.g. the bulk of your work is typing, and doesn’t require rapidly entering digits while using the mouse, such as a FPS shooter game would), then a keyboard like ZSA’s Planck EZ is hard to beat.

Those are quite expensive. For a lower cost option from AliExpress, the Idobao ID75 looks compelling. It fits in a GH60-sized case (which are easy to find), and like (I think) the above it’s got a powerful ARM MCU, hotswap (so no need to solder), and the fancy per-key RGB and RGB underglow.

For the smaller planck-size, previously I’ve used and recommended the BM40 keyboard. It’s still … okay. The main drawback to the BM40 is that its MCU is the weaker ATMega32U4. It’s powerful enough for a decently sophisticated keymap; but you’ll not really be able to push the keyboard to its full potential.

Albeit, I don’t use any of these since I got deep enough into the rabbit hole of mechanical keyboards to make my own DIY designs. I share these designs in this GitHub repository.


Newer post Older post