[microtonal]

As someone who has made a bunch of keyboards (including hand wiring), I agree that the space is very hard to navigate. Especially since /r/ErgoMechKeyboards/ has gone full "the only ergonomic keyboard is a hand-soldered keyboard that only 50 other people use", while for most people a prebuilt ergo keyboard would be fine and a large improvement over a $15 black slab.

What makes it worse is that most designs do no have any long-term group studies at all. E.g. Dactyl Manuform keyboards are quite popular, but many long-term users report thumb injuries from using Dactyl Manuforms. Who knows what injuries many of the custom designs can cause long-term? Most hobbyist keyboard designers are not trained medical professionals and basically don't have much of a clue what they are doing.

I would recommend people to start with prebuilt keyboards that have been around for a while and are pretty well-understood and in some cases had input from or testing by experts. Some examples:

- Fixed-split angled Microsoft keyboards (some models had extensive group studies, etc. Some models do break fairly quickly.)

- Logitech ERGO K860

- Kinesis Freestyle Edge RGB. This is traditional keyboard but with split and optional tenting (both of which have been shown to be good in various studies, be sure to get the tenting kit).

- Kinesis Advantage, goes all-in with ergonomic features and requires relearning typing a bit. Has been around in various forms for 30 years and is well-understood. Most people who use one swear by it.

- Keyboardio Model 100. The precursor (Model 01) has been around for many years. The design seems sane and I haven't seen many complaints.

- ZSA Ergodox EZ/Moonlander: these are quite popular. But be careful, quite many people report having thumb issues after a while.

I don't know any proven designs with trackballs, I have seen some people complain about integrated trackballs becoming uncomfortable after a while.

[precompute]

There's no one-size-fits-all ergo keyboard, which is why the dactyl has a configurator. You need to make test prints for your hand / finger / thumb size and splay and decide accordingly. Same for the other non-bowl keyboards like corne/lily58/etc, some people just can't use them.

I've been using a corne for two years with absolutely no pain/issues (not that I had any issues before I started using it, though)

[microtonal]

There's no one-size-fits-all ergo keyboard, which is why the dactyl has a configurator.

This is a bit of a trope in places like /r/ErgoMechKeyboards/. However, it's a bit of an empty statement without further data. Of course hand sizes/shapes differ, but the question is if they differ enough to preclude designs that work for a majority of the population. If you look at e.g. hand length/width datasets, within a group (e.g. men) these are close to a normal distribution. And the spread is not that large. e.g. the different in hand width in the ANSUR II 2012 dataset is 0.6cm between the 25th and the 75th percentile for men. The difference in hand length is 1.3 cm between the 25th and 75th percentile.

So, it's very likely to be possible to design a keyboard that fits most people well enough. Sure, if your hand size is an outlier you may have to get creative.

Besides that, there is also some latitude in designing ergo keyboards for multiple hand sizes. See eg. how Moergo iterated for several years so come up with a design that works for many different hand sizes [1].

[1] https://www.moergo.com/pages/glove80-ergonomic-keyboard-desi...

[precompute]

Hand sizes differ a lot, and finger mobility does too. Also, the length of the ring finger is different for different people. So yeah, there's a definite requirement for different keyboards for different people. And it's not a "meme" or a trope, it does make a huge difference. Plus, it depends on the layout you use (or are hoping to build) and the work that you do with your keyboard, and the kinds of features you're using (tap dance, macros, expansion, combos, etc) and how many of them you can remember. Sure, you can go with a bog-standard split and it'll be MUCH better than a regular QWERTY. But if your needs dictate a better keyboard, it doesn't mean you're hyperoptimizing.

[CarVac]

Hand width and size are not the only things to take into account.

The resting curl of your hand is important, as is relative finger lengths (particularly of the pinky).

And then there's resting thumb position relative to the fingers.

[microtonal]

The same applies to all these dimensions. They are not completely random and most will be normally distributed when measured. Also, as I said keyboards can accommodate multiple lenghts e.g. by having multiple thumb keys where you use a subset of them.

At any rate, I have seen plenty of folks who had issues like wrist pains, etc. that were resolved by improving posture, taking regular breaks, doing exercises and switching to a split, tented keyboard (just a boring one with a traditional layout like a Freestyle).

Only people who went (too?) far down the rabbit hole do all these micro-optimizations and proclaim that their keyboard is 'the end game' after having used it for a few weeks. Until the next end game :).

[CarVac]

They may be normally distributed, but most people have at least one or two outlier parameters that customization could help with.

Personally I just suggest going to a keyboard meetup and trying out a bunch of ergos and picking the design you find most comfortable.

I lucked into the Mitosis layout in 2017 (literally, I won it in a raffle) and have used it ever since. True endgame.

[loriverkutya]

Could you show the data this statement is based on?

|most people have at least one or two outlier parameters that customization could |help with.

[leononame]

I'm really looking for a good ergonomic keyboard right now (split or non-split). I'm using Redox, but the bamboo case is falling apart and sometimes I miss F-keys (it'd be nice to one-shot steps during debugging rather than having to press a key combination) or a physical numpad (for entering numbers I prefer the numpad, but I just don't love switching layers).

The Ergo K860 looks nice. I've also been eyeing the Kinesis Advantage 360, though it has no F-Keys. Another option would be the Glove80. The MS/Logitech keyboards are super cheap in comparison. Would you say not having ortholinear keys is a big difference for ergonomics?

[microtonal]

Would you say not having ortholinear keys is a big difference for ergonomics?

I don't think there is much proof in terms of scientific studies that column stagger is better (column stagger is usually preferable over ortholinear), so can only really give my personal experience. I strongly prefer column stagger, I can still type row stagger and sometimes use my laptop's keyboard when I travel, but all these lateral movements feel awkward now.

I've also been eyeing the Kinesis Advantage 360

The 360 is really nice (I have a 360 Pro), but I'd order it somewhere where you can return it easily (eg. Kinesis themselves). For me, the Gateron Browns in the 360 are quite a large regression compared to the Cherry Browns in the Advantage2. The Gaterons are smoother, but their actuation point is way lower. Gateron specs say 2mm, but it's closer to 3mm in practice. This makes typing on the 360 quite tiring for me. So it's mostly collecting dust currently (I am using the Advantage2 and the Model 100 most of the time). But this is really personal, some other people do like the Gateron switches.

I wrote up a short review of the 360 Pro here:

https://danieldk.eu/Posts/2023-04-15-Advantage-360-Pro.html

[goosedragons]

I really like my regular Kinesis Advantage 2 which does have F-keys although admittedly small ones. I don't use them too much. They are easy to reach which I like though.

I think having Ortho layout makes using the numpad on the Kinesis a lot better. The big ergo benefit to me is the thumb clusters and the fact I can remap keys like (),[],{},~ to be easier to press. I have setup the "delete" key on the thumb cluster to be a layer shift and moved symbols to easier to press fingers than ridiculous pinky extensions. The numpad is also easy to use since I just hit delete and then it's already under my right hand and I don't have to move it. Since it's Ortho it feels like a numpad. Previously I never really used the numpad because I didn't like moving my hand over but now most of my numbers are typed that way.

I don't think MS/Logitech boards go far enough even in the split department.

[cduzz]

I've found that the best "split" keyboard is simply to use two decent keyboards on the same computer. When I am in a situation where I'm doing tons of typing (not that often anymore) I'll have the two keyboards a bit more than a body width apart. Unfortunately macos doesn't apply modifier keys (shift / meta / ctl) from one device to the other (IE if I press left shift on the left keyboard I don't get a capital L if I press it on the right keyboard). Windows did this just fine.

For just general sysadmining around, one keyboard's been fine for me. Frequent breaks, standing up and moving, etc, have much more of an impact on fatigue than the quality of the keyboard. That said, I've found that the 1990s model M4-1 to be an absolutely spectacular keyboard.

[goosedragons]

Can you use something like Karabiner Elements to fuse the two keyboards together into one virtual one? I do something similar on Linux with mice for middle mouse button scrolling since my Ploopy Nano doesn't have buttons.

[cduzz]

In my youth I'd endlessly try to optimize things. You should see my fvwm setup. "Chef's kiss" perfection.

Regarding 2 keyboards -- mostly it just works. Try it. If you're like me you probably have a stack of random garbage keyboards within 30 feet of wherever you've got your work desk. Plug 'em in and type away. For me, anyhow, it works just fine - my left hand doesn't actually need to be within a foot of my right hand.

[microtonal]

I don't think MS/Logitech boards go far enough even in the split department.

I agree that the split of the Advantage is nicer. However, the MS/Logitech boards do not only split, they also put the halves at an angle, compensating quite a bit for their relatively narrow split. There are some papers that investigate these compromises.

In the end it depends on the amount that you want to spend. A 420 Euro Advantage2 is probably going to be better. But the Microsoft/Logitech/Alice boards are a good compromise if you do not want to spend more than 100-200 Euro on a keyboard. Though at the 200 price point it's probably better to stretch it a little and get something split and tented like the Freestyle Edge, which sometimes goes for 200 + 30 for the tenting set from some vendors.

(Sometimes you get lucky, I found a mint Keyboardio Model 100 on a local marketplace for 150 Euro with the octo feet. It's over 500 Euro after VAT and shipping when you import it to The Netherlands.)

[lars_francke]

Thank you very much for the detailed response. I'll look at all those options. I've been using a Microsoft Natural Ergonomic Keyboard 4000 for ages and it is slowly disintegrating so I'll need a replacement in the next few years.

The quality has been great. It's lasted 15 or so years now.

I switched to an optimized keyboard layout (neo2 - German) in 2006 when I had carpal tunnel syndrome in both hands and then also switched to one of those ergonomic keyboards.

Now, I'm basically looking for a successor.

[coldpie]

Check out the Feker Alice 98. It just came out (my pre-order arrived just last Friday) and is an excellent modern recreation of my favorite keyboard, the Microsoft Natural line of keyboards from circa 2000. There's a separate model without the numpad, if you don't use that (I do).

Complaints? No Home/End. The extra B on the right side is weird. The proximity of the arrow keys means sometimes I hit the left arrow-key when pressing backspace. But those are all easy to adjust to, I think. I'm excited to finally have a good, modern replacement for the Natural keyboards. Highly recommend this one.

[tom_]

My ~2000-era MS Natural keyboard has the following. (I keep it in a drawer, as an emergency spare for when I run out of MS Natural 4000s.)

- Full navigation cluster

- Arrow keys away from the main area

- Ctrl+Shift in line on both sides

- Right-hand Alt key

- Right-hand Windows key and right-hand Windows context key (you don't need both IMO, however having one is useful as you can remap it to be the right-hand version of the one you do use - the MS Natural 4000 is fine)

- Only one B key

- f6 on the left side of the split

For a device that relies so much on muscle memory, you'd think that designers would do a better job of laying theirs out exactly like every other one. But... apparently not?

My search for a replacement continues.

[coldpie]

> My ~2000-era MS Natural keyboard has the following. (I keep it in a drawer, as an emergency spare for when I run out of MS Natural 4000s.)

Hahaha! I have the same thing. I've got one original MS Natural still BNIB, and two Elite models each with a lot of mileage on them. I've long been dreading the day that my stash runs dry. Been keeping an eye on keyboard forums for years waiting for a replacement. The popularity of high-end keyboards with the terrible rectangle layout utterly baffles me. Why would someone who cares about keyboards ever choose that layout?

The Alice layout comes close enough to the Natural layout that I think it's a fantastic replacement. My complaints are small enough that I'm already adjusting after just a few hours of usage. I'm just thrilled someone is making a split ergo keyboard with a numpad!!

[tom_]

Indeed. The weird navigation keys are weird, as they're in funny places and half of them are missing. And I might have some suggestions about better places to put the arrow keys. But there's a numeric keypad, and there's a full set of F keys! Progress!

[robohoe]

How about a split mech like Mister Barroco? Easy to find in online shops. It splits and you can have a trackball in the middle.

[comfypotato]

Can you elaborate on the thumb problems from Manuforms? This is the first I’ve heard of them. I’m in the process of building a corne Manuform.

[microtonal]

https://www.youtube.com/watch?v=foSmpE_aefI

And basically a lot of reports on Reddit. Eg.:

https://old.reddit.com/r/ErgoMechKeyboards/comments/ubcxup/e... https://old.reddit.com/r/ErgoMechKeyboards/comments/umg2j5/e... https://old.reddit.com/r/ErgoMechKeyboards/comments/lj3n45/p...

Of course, these are just anecdotes and we don't have good research into all the individual designs, but there seems to be a pattern. The movement is also similar to what causes texter's thumb.

At any rate, I am not an ergonomics professional, so if you are switching to a more ergonomic keyboard because you have pains, you should ask and expert what works and what doesn't.

[comfypotato]

Thanks!

You’re really on one on this thread lol.

I’ve found that anecdote gets the job done for me. The $ values are insignificant for me considering it’s all work-related. The RSI has come and gone through multiple iterations of work intensity and keyboard (and accessory) configurations.

I suspect this is the main reason there isn’t much research in the space. The breadth is too large to research when trial and error quickly arrives at a workable solution for 99% of people.

With a little bit of technical insight into RSI, common sense goes a long way into finding tech configurations that work.