Comment by eviks
2 days ago
The progression from unergonomic garbage default to splits and ~glovey form makes sense (with the future more sensitive per-finger sensors allowing for more gestures?), though it's a pity the health issues were involved, and that the resulting niche is small and expensive. Where are all the cheep rubber dome splits? Those would already offer health advantage without the mechanical premium...
By the way, the layout might better use physical direction mnemonics, so backspace could be a finger move to the left, and delete - move to the right (either by the same finger or maybe by its mirrored counterpart on another hand). Similar thing for <> and -+-+
> north keys are much harder to hit than its south keys, > find the inward lateral keys much easier to hit than the north keys
So, basically, "grabbing" movements are more natural?
The think the premium that these devices suffer from is how niche they are. Mechanical is relatively niche to begin with (programmers and gamers, mostly). Then you add the idea of making a single large investment for the good of your health, and the pool gets even thinner. Just think of all the garbage chairs that people sit on in front of their computer; let alone keyboards, which they barely think of (until they get RSI, that is).
I'm currently designing one that is a crossover between the characorder and a Logitech vertical mouse (but not actually a mouse), and I hope to make it cheap. It has been going poorly, to say the least. The biggest problem so far is smaller hands, multiple designs scrapped when put in front of such a person. Maybe I should take inspiration from Svalboard instead? Furthermore, Average Joe is not going to tolerate the massive dependency on layers that these keyboards often have, so I'm still fighting in my head with an approach for that.
The Datahand / Svalboard branch is even more niche than a standard mech board. They are really designed for people who are facing injury, injured, or have been injured.
I'd encourage you to look at the Svalboard, I think it solves many of the issues you are running into. In the end, moving a large mouse is not good for someone with shoulder issues, the "Svalmouse" has been tried and solidly rejected. (Using a Svalboard hand as a mouse.)
If you want to discuss things, I'm sure you can find me on the Svalboard discord. :)
I'm not sure you understood me: it looks like the ergo mouse (form factor, ergonomics) but is most definitely not a mouse. As I said though, it seems more difficult to get a keyboard to fit a variety of hands in this form-factor.
2 replies →
That sounds really cool! I'd love to see it when you've got something more concrete. I assume you're already aware of the KeyMouse and the AlphaGrip?
https://www.keymouse.com/ https://alphagrip.com/
I don't think the layering needs to be as big a deal as one might assume. I mentioned in the post, but in many ways I think layering is easier to deal with than larger keyboards, and not only that but also culturally small-keyboard users are increasingly okay with using layers, e.g. the tiny-keyboard gamer crowd, or even most laptops now have a dedicated Fn layer in addition to the traditional shift, ctrl, and alt. So long as you don't go overboard with it I think it shouldn't be that intimidating.
I grabbed an alphagrip from the ewaste bin at my local hackspace a few years ago. Gave it a solid week - which was enough for me to get at least marginally competent with the layout. I found it extremely awkward. I think the compact controller style forces your wrists into an awkward angle, so any advantages of minimizing fingers motion is compromised by the awkward neutral position. Sloppy back switch mechanisms and a really bad trackball on top.
Too uncomfortable to use full time, to awkward to hunt and peck for an occasional couch/TV navigation keyboard - It went back in the ewaste bin.
I've been on a kinesis advantage 360 for a few years, after an ergodox and the older fixed size Advantages. Been eyeing the Svalboard, thanks for sharing your experiences!
1 reply →
More things to ~steal~ take inspiration from, thank you :). If I'm completely naive with my approach, I would have 5 fingers * 4 directions * 2 hands. That's a paltry 40 keys. To make matters worse, some of them are sub-optimal: the ring finger isn't great at sideways movements (at least on my hand). 35-40 keys is end-game enthusiast level minimalism.
> Where are all the cheep rubber dome splits?
The only halfway decent player in this category was Microsoft, with its line of Natural Keyboards. I've used four or five of them, decent enough. I doubt Microsoft was making a lot of money.
The line was discontinued in 2023 and sold to Incase:
https://www.incase.com/pages/incase-designed-by-microsoft-co...
Kinesis now has https://kinesis-ergo.com/keyboards/mwave/ as a better(to me) Microsoft sculpt clone in a similar price range.
Looks good, they kept the negative tilt which I find very important but nearly impossible to pitch to non-ergonomic keyboard users.
I've been pretty happy with this as a MS 4000 replacement: https://eu.perixx.com/products/periboard-535 - general tent/reverse tilt/split angle feels about the same. Like the MS 4000, it's a full size keyboard, with symmetrical sets of meta keys, unintrusive level of F-lock nonsense, easily accessible keyboard volume controls, and a generally standard layout that will probably require only minimal retraining. Also connects to your computer using a cable... none of that wireless nonsense. The UK layout version is under £100 on Amazon.
It feels a bit less solid than the MS 4000, but it's nicer to type on.
(The MS 4000 was always a huge pain to repair (endless screws to get inside, many annoyingly inaccessible; never anything obviously wrong once in there; time-consuming cleaning/drying-out process; endless screws to put it back together again), but this looks like it might be a bit better. Not many screws on the bottom. It isn't thick enough for there to be any hiding inside. The keyswitches will probably be individually replaceable. Ask me again in 10 years though!)
I agree, I never wanted to get into mechanical keyboards per se — I was always just on the lookout for an ergonomic layout. Mechanical switches are cheaper and cheaper these days, though, and I suspect the premium on high-end keyboards now is more from a combination of the luxury good tax and the lack of economies of scale.
I really wanted to make myself a steno theory for the CharaChorder, but now I don't have time. There's definitely something to be done with keyboards like this and I do reckon you could do better than with the mechanical constraints of the stenotype, but it's not trivial: the extra inputs are hard to make use of if you want to avoid awkward ‘scissoring’ strokes, though there are definitely some comfortable strokes in there to mine.
The physical direction thing is a neat idea but a bit hindered by the fact that the two sides have very different accessibility: typically the inward keypress is much easier than the outward keypress. There might be some pairs that's good for, though, e.g. the backspace/delete pair you mention (if put on the right hand: delete is much less common, in my experience, than backspace).
> So, basically, "grabbing" movements are more natural?
Right, for me at least that seems to be the case. Other Svalboard users seem to pretty unanimously agree that outward movements are worse than inward movements, especially on e.g. the ring finger, but people differ on how the inward/upward/downward movements compare.
> typically the inward keypress is much easier than the outward keypress.
That's why my default suggested mode is to mirror, for example:
Or right middle finger down for > and left middle finger down for <
Right, that's actually what I do! Check out my symbol layer: https://twey.io/hci/svalboard/layout-symbols.png
Disclaimer: Volunteer Svalboard Firmware Dev.
For most people the south keys and center are easier to press and more importantly hold. I don't care much about directions anymore, except for holding down keys.
Layout is a personal thing, we use Vial, so different people have different layouts, mine has a ton of mnemonics in it. It's really up to the user what they want, I'd be surprised if there are 2 Svalboards with the same layout. (Maybe stock when they are starting out, but people develop their own opinions pretty quickly.)
Keychron has some "ergonomic" mechanical keyboards that are fairly affordable. It's not perfect (no tent, no split), but overall it's a nice compromise.
Whoever is first to mass-produce a cheap rubber-dome split columnar keyboard is going to sell a lot of keyboards, and kick the ergo keyboard revolution into high gear. Just imagine selling 90% of a $349 keyboard.io Model 100 for $20, and it only cost you $3 to make. Throw it on the shelves and watch it sell itself. It's a novelty. It's an accessibility tool. It's the first step into a new hobby. You wouldn't even need to advertise; people would buy it just because it's interesting.
I finally made it into a Micro Center yesterday. There isn't one anywhere near my home state, so I jumped at the opportunity. What did I buy? Two ortholinear keyboards. There is a huge market for these things, and even the geekiest big-box store in the nation only bothers to sell 2. Why? Because nobody wants to spend more than $30 on something they haven't ever tried.