> One week or so ago I switched from Colemak to typing on QWYRFM due to the reduced load it places on the pinkies.
I typed Dvorak for 10 years before I started getting RSI again in my right ulna. It turns out that the 'S' key being hit by the pinky in Dvorak caused me a lot of pain.
I ended up switching to BEAKL, a pretty obscure layout which is specifically made to alleviate stress on the pinky. It moves the relatively common S key to the right hand pointer finger, for instance.
It turns out that the theory underpinning the dvorak layout assumes that homerow hits are strictly better than upper and lower row, but the pinky and ring finger are such weak fingers that it's often better to move the index or middle fingers up or down a row than hit a key on the home row with the pinky.
A friend of mine used his pinky and ring finger to open the door of his pickup truck. But it was frozen shut and he snapped the tendons (ligaments?) and that was that, he couldn't curl them shut on his hand anymore.
What? Did he not go to the doctor? That's a fairly straightforward surgery to repair even though the recovery and therapy sucks.
It's way easier to repair a "snap" than a "slice". However, you do have to get to the surgeon before 14 days are up because everything starts shrinking immediately. If you don't, then they have to graft a piece of vestigial tendon from your arm (apparently about 85% of people have it) to make up the length.
Surely that heals after a few months? I tried opening our shower door which magnetically attracts and seals the glass doors together with my pinky, I wasn't prepared for the amount of force required to open so I got sharp pain and had problems doing anything with my finger. Luckily it only lasted a day. Reading this I didn't think it could've gone this bad.
> but the pinky and ring finger are such weak fingers that it's often better to move the index or middle fingers up or down a row than hit a key on the home row with the pinky.
Spot on, I would go as far as saying that using the pinky at all is poor technique because it indicates one reaches keys by stretching the fingers, rather than hovering the hand over the keyboard, allowing each finger to remain in its natural position, and using only the fingers that are naturally strong and precise.
That’s interesting. I guess I never really thought about it, but on dvorak `ls` is a double tap with the right pinky, and it’s often immediately followed by [enter] with the same pinky… I type that key sequence a lot
Highly recommend you look into keyboards that have "Enter" and "Backspace" reachable by thumb/index fingers. I've been dailying a TECK for nearly a decade, and it's helped immensely with RSI.
Their new Clevo keyboard is a bit different, so can't vouch for it, but my SO likes it so far as first foray into ergo keyboards: https://trulyergonomic.com/
I would really want to try that, but it seems extremely tied to the language. for English that's cool, but I write code, greeklish (Greek with English characters) and English daily. That means that for these three I would need to configure things and that seems like waaaaay to high of an up front investment for a trial.
I used some generation of Truly Ergonomic for a few years and it started getting repeating/missing characters... at $250 it was too big of an expense for me to rebuy. While it worked though, I was very happy with it
My impression was that they used some slightly chattery Cherry MX keyswitches combined with a controller with a poor debounce routine. (The electrical output of many switches "bounces" quickly on and off a few times when changing states. This can cause dropped or doubled key presses if not accounted for by the keyboard controller.) I’m not sure if the batches of Cherry MX switches they had were bad, or if some amount of chatter is inevitable given the switch design.
You should try the ergodox. I similarly bought the TECK because I wanted an ergonomic keyboard with a standard-ish layout and loved it, but it died in a couple of years. After that I bought an Ergodox EZ and have been very happy with it.
to my .zshrc about 10 minutes after I started using Dvorak, 18 years ago. "ls" on its own is OK. It's not an especially common bigram (wails, fills, also, Welsh, whilst), but I can't find any English word matching ls[rlvz], which is the pattern of "ls -l". Compulsive, compulsory, falser, impulsive, maelstrom, pulsar, wellspring come closest.
I believe LSZ is all with the pinky in the usual method[1]. There's no other option on the keyboard I bought a couple of years ago, which also has several thumb keys.
I fixed my one case of budding RSI with Two Weird Tricks.
The important one was to make my home row on laptop and other 'standard' keyboard layouts into (approximately) QSDV-POKJ. That is to say, my palms are angled relative to the keyboard angle. Without this, the wrists are squished the whole time one is typing: that's bad and no layout can fix it, it's pure posture.
The other one was to move backspace to the cap's lock position. Turns out I delete a lot and i was doing it with a rather extreme wrist rotation and my pinky finger.
I use an ErgoDox when I can these days, but can easily spend an hour or two typing on a laptop keyboard, my spine needs a break well before my wrists would.
Interesting. I use Dvorak too and find it much easier than qwerty. But I have been wondering about that right pinky. I broke mine when I was much younger (skulls are harder than knuckles, kids). I guess I will try your BEAKL suggestion.
Thanks for sharing this. As a Dvorak typist of 6 years, I have often wondered what my next frontier will be. Stenography seems to have a fast growing community around it, but I'm open to alternative layouts as well. BEAKL has always been on the edge of my radar holding the position of interesting, but probably no. This however is the first time I've seen someone compare it favorably to Dvorak. Seems it could be worth trying out based on what you said! While I feel like Dvorak has strengthened my right pinky finger, it's L, /, and = keys are notably tedious. I wouldn't want to have RSI as a result of that.
Did you ever try Colemak? I've been Colemak for probably 5 years now, I can't speak to pinkie specific pain but I get much less tired on long typing sessions. It's interesting to me how everyone seems to know of Dvorak over Colemak. But I never heard of BEAKL, thanks. For me the issue with Dvorak was that I wasn't ready to give up my function keys (control x, c).
I haven't, just went straight to dvorak. Afaik Colemak has less intense right pinky usage but more same-hand rolls.
Losing the function keys is a pain, on MacOS you can use Dvorak-Qwerty which retains them, on my new keyboard I just have a new "command" layer which retains the old positions.
I used BEAKL as well for a few months but in combination with a 5-column split keyboard (so you don't press enter, shift etc with your pinky) there was too little for my pinky to do (which caused other fingers to do more than needed).
Of others have similar issues I would recommend you to look at a split keyboard first (but switching layout is great too).
That's interesting, I do type on split ortholinear keyboards (e.g. corne variant) so maybe I'll run into that soon. My right hand pinky is pretty bad with flare-ups so it would be a welcome break for me
Interesting. I've never trained or practiced my typing, but I've always naturally typed using only my thumb, pointer and middle finger, no pinky or ring fingers at all, but still manage to hit 80-100WPM without them on QWERTY. Good instincts I guess.
So small caveats, I lied a little. This is the first time I've ever really sat down and analyzed my typing. I do use my ring fingers slightly, on my left hand I use my ring finger for "z" "a" and "q" exclusively, and on my right hand I use the ring finger for backspace exclusively. I also only use my left thumb to hit the space bar and use my right index finger for space when my left thumb isn't available, and don't use my right thumb at all.
I managed to hit 93WPM(really 102WPM if you look, it wouldn't continue until I finished the last word which I didn't notice for quite a few seconds lowering the final time) my first try while holding my phone in my mouth which was a little distracting to be honest, and after about 10 minutes of warming up and not holding the phone managed to get it to 132WPM.
First try at 93WPM - https://gfycat.com/groundedpettyeider
Had to cut it to 60 seconds to fit into gfycat, so it starts a little into me already typing but gets the idea across
There was some research that showed people who typed with fewer fingers could be just as fast. However, I am surprised that anyone could achieve 100wpm with no training.
I've never done a ten finger / touch typing course or something like that (so, no training) and don't have problems typing 130 WPM 98-100% accuracy in one of these typing tests without specifically practicing typing tests. I don't have ten working fingers.
So small caveats, I lied a little. This is the first time I've ever really sat down and analyzed my typing. I do use my ring fingers slightly, on my left hand I use my ring finger for "z" "a" and "q" exclusively, and on my right hand I use the ring finger for backspace exclusively. I also only use my left thumb to hit the space bar and use my right index finger for space when my left thumb isn't available, and don't use my right thumb at all.
I managed to hit 93WPM(really 102WPM if you look, it wouldn't continue until I finished the last word which I didn't notice for quite a few seconds lowering the final time) my first try while holding my phone in my mouth which was a little distracting to be honest, and after about 10 minutes of warming up and not holding the phone managed to get it to 132WPM.
First try at 93WPM - https://gfycat.com/groundedpettyeider
Had to cut it to 60 seconds to fit into gfycat, so it starts a little into me already typing but gets the idea across
I just use query, but I only type with 3 fingers on each hand and move my hand if I need to. I've always felt holding my hands rigidly over the keyboard was going to give me rsi no matter what the layout.
I self taught myself to type and so don't really follow whatever is "proper" technique, so with WASD I tend to use my left pinky for SHIFT and nothing else.
This is surprisingly common. WASD-biased finger positioning. Also if you only use left shift it reinforces the habit since your pinky is what would normally press the a key. If you ever fix your typing technique (or just get curious), ESDF (instead of WASD) kinda fixes this. Your same fingers handle movement, but while also using proper homerow, and you free up space left on the movement keys where you could move stuff that was on ctrl and shift to the letters. (talking about the in-game key bindings here)
A few years ago I relearned qwerty with proper homerow, started using right shift, etc. I then later switched to Dvorak and then Workman. If I ever play a game I now set up the equivalent to qwerty ESDF in the controls, but with the letters of my layout. I sometimes also set backspace to jump instead of space because I use a split keyboard where I placed my space key on the right half thumb cluster and my right hand is on the mouse for games like Minetest and Xonotic.
I switched to Dvorak in my 20's. I don't think i got faster, but my right-hand wrist which would irritate me and was one reason for the switch, definitely was more comfortable after, and hasn't given me any problems since.
That said, switching was a huge pain in the ass.... like horrible for the first two months, and i promised myself i'd keep typing in both. However, i eventually lost the ability to type in qwerty, after about two years, but i honestly don't care, i genuinely love Dvorak.
The biggest concern most people probably don't think about is keeping shortcuts like CTRL+X/C/V/S/F, which apple thankfully has a layout for "Dvorak-CMD-Qwerty." The shortcuts are very obviously put where they are so you can toggle them quickly. Many programs don't account for this layout, which is why I wrote an Intellij IDE key mapping for it: https://github.com/scoofy/Intellij-IDEA-Dvorak-QwertyCMD-key...
The layouts that i think don't get enough love, are the one-handed dvorak layouts designed for amputees.
The idea that, in the end, the author suggests yet another standard seems a bit absurd to me.
I switched to Dvorak from QWERTY without any special control/cmd mapping. As it turns out, it works out fine. I even vim'ed with it. It's just finger memory. The hjkl all in a row might help you remember/initially learn, but once you know which keys on Dvorak move you the right direction, it doesn't take any more time or concern you any more than hjkl on QWERTY do. You know which direction you want to move, and your fingers hit the keys to make it happen without respect to them not having any meaning by keyboard position.
I switched after developing hand injuries at work (in a job that didn't involve computer use). Since then, whenever I've had to use QWERTY for whatever purpose I've started getting pain in my tendons after a few hours which goes away almost as soon as I'm able to switch back to Dvorak. Like you, though, I found that picking up Dvorak was a total pain in the ass. A while ago I saw a YouTube video in which a guy was telling his viewers that learning it was easy and that the effort was a small price to pay. When I commented that I had had a period of months when I couldn't type anything spontaneously, he suggested that maybe this was due to my age. I was all of thirty at the time I learned it.
Data point. Old fart who switched to Dvorak eons ago, at age 16. It felt natural after 1 week, and felt more efficient than the ‘qwert sometime between weeks 2 and 3. I suspect young age greatly affects take-up speed. Good on ya for taking care of yourself, and switching when you did.
Incidentally, at age 26, I used the “left-handed” Dvorak layout for 2 weeks while my right wrist and hand healed from a dirtbike crash. I was only becoming proficient at the 2 week mark.
The Dvorak left- and right-hand layouts are not as widely known, and look nothing like the standard two-hand layout… but are incredibly optimized. Even the number row is remapped to accommodate letters. Once I started to grok its efficiency, it felt like 10-key’ing entire words at a time. Might be interesting on modern phones.
I experience the same thing with QWERTY and tendons in wrist & forearm. I have to slow down - a lot - and take breaks. I just remap teh kbd if > 15 minutes.
iOS kbd keeps my brain familiar with QWERTY; no first-party Dvorak and I don’t expect Apple to ever include one.
I tried Dvorak on Android eight or ten years ago and realized that my Dvorak skill would not transfer to a touchscreen, since it consisted entirely of muscle memory and I had no experience of looking at a keyboard that was labeled that way. That and the lack of glide typing (which I've come to rely on) made it useless to me.
I've used dvorak and workman on Android, and although you don't exactly get perfect skill transfer, I like these layouts I've invested time in, I like the reminder that I put in that work, I like where the letters are. I am writing this very comment on my phone in workman. The consistency is nice, even if it's not consistency in something as useful as the tactile feel.
On the less-bright side, lack of other layouts on the PinePhone have made using it a bit torturous for anything extensive, and that makes me enjoy using mine for anything a bit less.
Took me about three weeks when I switched cold turkey at age 14 to start typing ok. My biggest pain was that the letter I search for was always under the other hand which was resting on the keyboard :)
I'm letting my daughter learn it when she's 8 instead.. they just.. learn it. It is truly fascinating from the perspective of a typist.
So maybe the youtuber is onto something. I think that you should have a clear strategy when learning though, that will probably speed up the process alot.
I've switched to Dvorak twice, first in my teens and again in my 20s. I'd never been great at typing with Qwerty (a lot of hunt and peck) so it was a low bar. I can absolutely relate to the terrible two months (at least) but I at least retained what little ability I had in Qwerty.
Both times I've switched back after a few years because of needing to share computers with other people and not wanting to deal with switching the layout (either physically or in the OS). I also had trouble with vintage computers and phone keyboard layouts which couldn't be switched. For some time I was faster at Dvorak and I really enjoyed that, but I think learning Dvorak and switching back has at least made me better at Qwerty. I love Dvorak and will almost certainly start using it again someday. Right now I've accepted I just need to go with the flow and do what everyone else is doing.
I've never had issues with typing. It's the mouse that irritates my wrist. I swap it from side to side to alleviate the problem. i do have wider hands...
I tried one of those, but you can only use them with left or right hand and the wheel still aggravates my wrist. Swapping from left to right seems better. I also add in a little weight training and I don't have any problems.
Did you check finger operated trackballs? They are symetrical and can be operated by both hands.
Examples: Kensington Orbit, Kensington Expert Trackball, GameBall trackball, Logitech TrackMan Marble, X-keys L-Trac.
I tried it too and lost qwerty skills. That and the shortcuts are the main reason I gave up. I work with other people and was surprised by how often I type on other people's keyboards. A nightmare if you mainly use Dvorak.
My initial motivation was also wrist pain. Dvorak helped a bit. I also tried crazy split vertical keyboards which didn't help at all.
What fixed it in the end was a bigger desk and a good chair (I have a HM Mirra 2 now; worth the extortionate price). You need your keyboard far enough onto the desk that you can properly rest your forearm on it.
> In other words, the idea is that QWERTY was designed so that humans would type as slowly as possible in order to prevent mechanical jamming.
Nope. That's an oft cited story but it's not true. The point of QWERTY was actually to increase typing speed. The debunking of this is pretty well documented these days (there even a note on wikipedia: https://en.wikipedia.org/wiki/QWERTY#Properties) so it's disappointing an article as researched as the one submitted here has made such an easy to check mistake.
But I think the bigger trap it falls into, and the same trap many on HN fall into when posting their anecdotes too, is that RSI isn't universally equivalent. It depends on hand sizes, the keyboards you use, any other hand exercises one might do, underlying health conditions, and even just how you hold your hands at the keyboard. So what might help alleviate RSI for one person could easily make it much worse for another person.
It does, but the article is pretty cagey in its retraction (including following up with the comment "Though on the other hand, jamming is really bad for typing speed!").
The author seems to me to at least partly still believe its true, merely acknowledging that it's a difficult position to maintain.
i am the author and i didn't mean to be wishy-washy about it. i read several of the papers involved and think it's more likely than not that qwerty was designed for fast typing, or at least not designed to type slowly. but i wouldn't say it's a done deal -- the paper i cited gives the most convincing argument, imo, but it didn't move me up to 100% confidence or anything. i didn't want to give the appearance of being too certain, because i'm not.
The confusion in many articles on the subject is because both the myth and the truth argue that QWERTY was designed to prevent hammer jams. The reason this confuses people is because anecdotally many people find QWERTY slower to find keys ergo many laymans assume QWERTY was designed to slow typing down to avoid jams (in fairness, it's not an unreasonable piece of logical deduction even if it's not technically accurate).
The reality was that typists were trained to type on QWERTY so they weren't slow typing on it. What QWERTY was designed to do was move common follow on characters further away from each other so the hammers jammed less often thus allowing trained typists to type faster than they could have on an alphabetized layout.
Given that the person you are replying to is the author, I'm guessing he probably does not "partly still believe it's true." I'm going to bang on the "give people generous readings" drum here, too: "Jamming is really bad for typing speed" is not "pretty cagey," it's "a humorous comment." (Also, you know, technically correct.)
The point of the QWERTY layout was to prevent jamming; the prototype layout was roughly alphabetical on the keyboard and jammed constantly, so it was refined to fix that mechanical problem.
The part that is a myth was that they prevented jamming by “slowing typists down”. What they did is moved commonly adjacently pressed letters apart from each-other so they wouldn’t be pressed at about the same time. This was not done explicitly to speed or slow typists (at the time, there was no established typing technique and there were no expert typists), except insofar as a constantly jammed keyboard was extremely slow.
But the nearness that mattered was in the ring of typebars on the Sholes/Glidden/Schwalbach/Densmore typewriter, not the nearness of the keys on the keyboard per se. This typewriter connected the keys to typebars which were placed around a ring and struck upward at the platen. The association between keys on the keyboard and the positions around the ring is not obvious.
They started by trying to place common English digraphs onto typebars at opposite ends of the ring, then swapped keys around via trial/error until they got the jamming problem mostly under control. Later the Remington company did a few more swaps so that "TYPEWRITER" could be typed entirely on the top row, as a marketing gimmick.
The true part of the common QWERTY criticism is that the layout of the keys was not decided primarily based on the positions on the keyboard or human hand anatomy. As a result, the keyboard layout per se (both the positioning of the physical keys themselves and the choice of association between letters and keys) is not especially efficient or ergonomic.
A related (milder) criticism can be leveled at the Dvorak layout, which was designed for the typewriter technology of ~1930, where the home row and top row of the keyboard were easy to press but the bottom row was much slower and less convenient, based on the consistent steep step between rows of keys. Computer keyboards have always had a different physical shape compared to 1930s typewriters, so the relative difficulty of typing different key combinations is different, and an optimized layout therefore looks different.
> The part that is a myth was that they prevented jamming by “slowing typists down”. What they did is moved commonly adjacently pressed letters apart from each-other so they wouldn’t be pressed at about the same time. This was not done explicitly to speed or slow typists (at the time, there was no established typing technique and there were no expert typists), except insofar as a constantly jammed keyboard was extremely slow.
Indeed, which was designed to speed typing up, hence my point. :)
> The true part of the common QWERTY criticism is that the layout of the keys was not decided primarily based on the positions on the keyboard or human hand anatomy. As a result, the keyboard layout per se (both the positioning of the physical keys themselves and the choice of association between letters and keys) is not especially efficient or ergonomic.
I get that but it doesn't take away from my point that the RSI complaint is hugely specific to the individual. Or to put it another way, if QWERTY were really as bad as the stories posted on HN make out then you'd see more cases of RSI given it is used by literally hundreds of millions of people every day.
I'm in my 40s and have been using QWERTY since the 70s without any side effects (I did start getting RSI in the arm from playing FPS every waking moment back when PC 3D FPS were new, but never from typing).
> A related (milder) criticism can be leveled at the Dvorak layout, which was designed for the typewriter technology of ~1930, where the home row and top row of the keyboard were easy to press but the bottom row was much slower and less convenient, based on the consistent steep step between rows of keys. Computer keyboards have always had a different physical shape compared to 1930s typewriters, so the relative difficulty of typing different key combinations is different, and an optimized layout therefore looks different.
Which again also misses my point that RSI depends on vastly more than just your keyboard layout.
There are an awful lot of people who develop RSI from keyboard use, nearly all of which is preventable. Better training would help a lot, but better designed equipment would also make a big difference. Changing the QWERTY association between keys and letters shouldn’t be the highest priority, but the advantage it has is that it can be done in software, without change to physical devices. Reshaping the keyboard to better match human anatomy would make a bigger difference.
> There are an awful lot of people who develop RSI from keyboard use, nearly all of which is preventable.
I'm not saying there aren't. What I'm saying is it's not directly because of the keyboard layout. You go on to say this yourself in the very same post as your point above:
> Changing the QWERTY association between keys and letters shouldn’t be the highest priority,
People talk too much about keyboard layout with regards to RSI and totally blank out all the other factors that contribute.
If the problem was only related to the keyboard layout and QWERTY is as bad as HN commenters make out, then we'd hear a lot more cases of RSI. But it isn't, which statistically suggests the problem is more nuanced. ie type of keyboard (mechanical, chiclet, even down to the specific manufacturer and the parts they've used), position of keyboard in 3D space relative to the operator, position of hands on keyboard, typing style, health of the operator, operator's workflow, etc. There are so many variables at play that blaming the keyboard layout had always struck me as a massively lazy deduction.
I think you dramatically underestimate how many people suffer from RSI and how much it affects their lives. In the US alone, we are talking about millions of people. In some surveys >50% of keyboard workers have complained about wrist pain from typing.
There was some planned OSHA office worker ergonomics enforcement in the late 1990s, carefully studied for a decade before being finally put into place in 2001 just as Clinton was leaving office, but it got killed by GOP congress and the Bush II administration. If not for that we would hear more about this problem, and there would be a lot more options for ergonomic keyboards today. Instead, this became a problem that employers blamed on their workers and left workers to figure out for themselves.
> position of keyboard in 3D space relative to the operator, position of hands on keyboard, typing style, health of the operator, operator's workflow, etc.
Yes of course. And things were to some extent better when almost all of the people using keyboards were trained professional secretaries.
From walking around large offices full of computer programmers, I would speculate that at least 3/4 of current office workers have inappropriately positioned and oriented keyboards; maybe even 90% or more.
The biggest problem (among many) with modern equipment/practice is that people type with unstraight wrists, slouch or lean over in their chairs, and stick their elbows out forward or way out to the side.
The keyboard should be held close to the torso and angled so that its top is parallel to the forearms (what this angle should be depends on relative height of torso and table or keyboard tray). Typists should keep their wrists as straight as possible (not flexed or extended; pronation is unfortunately unavoidable on a standard keyboard), and should have their palms and wrists “floating” above the keyboard while actively typing. The primary goal should be to avoid any nontrivial static load on any muscles throughout the body, and try to keep joint positions neutral.
Splitting the keyboard in half, separating the two halves by at least a few inches, and independently positioning and orienting each half can make a huge improvement.
> [If] QWERTY is as bad as HN commenters make out [...]
There are very few HN commenters who care at all about QWERTY, and even most of those are relatively ambivalent. It’s a design poorly matched to modern hardware, and changing it can make a nontrivial improvement, but it’s also not the main problem.
> I think you dramatically underestimate how many people suffer from RSI and how much it affects their lives.
I'm not saying it isn't high; I said it would be higher. The distinction here is vital, as I'll go on to explain...
> According to one survey, nearly 60 percent of computer office workers nationwide suffer from wrist pain while at the computer, and 51.2 percent say their keyboards are placed too high.
So more than 5/6 people who have keyboard related pains believe it is due to the placement of the keyboard.
Of those remaining ~15% of users, how many of them are getting pains due to other circumstances (keyboard type, other positions in 3D space, or even just their own posture)? How many of them might still have their keyboard too high and don't even realize it (those excluding them from the first category where they should have belonged)?
Hence why I said the number of complaints would be much higher if QWERTY itself was a major contributor to RSI because QWERTY is near universal whereas the other factors are not yet the other factors still vastly dominate the statistics.
As for the rest of your comment, you're literally just reiterating everything that I've already said to you in my past comments.
To summarize (because I think you're fixating on the detail and thus missing the overview for it): for all the talk about how "bad" QWERTY is, other environmental factors have a much more significant impact on the health of the typists.
The story about marketing and "TYPEWRITER" on the same row is also considered a myth by some historians, simply because Remington spelled it "Type-Writer" and - was not on that row.
That the layout was intended for typing at speed is corroborated by the fact that early users of (pre-Remington) typewriters were telegraphists who used them to transcribe telegrams from Morse code at the speed at which it was being received.
Fair enough about the Remington 'TYPEWRITER' story; I don’t have direct evidence there, only what I have read elsewhere (which could be apocryphal). Someone at Remington might have swapped R with . because they wanted to keep letters separate from punctuation, or the like.
Anecdotally, some years back I started developing pretty bad RSI from typing. I was also a pretty mediocre typist. So I switched to Dvorak and got serious about touch-typing. I found it much more comfortable and since then I've never had any problems with typing-induced RSI.
With QWERTY, I was doing all sorts of things to try to stop the RSI. I tried to control the position of my wrists, the angle of my hands, how I hit the keys. I experimented with wrist rests. Since I switched to Dvorak I found I didn't need to pay attention to any of that. I can type virtually continuously for hours at a time at >100wpm, and while my speed will drop from my hands and the muscles in my forearms getting tired, even at that point I don't get RSI.
I had another set of RSI not that long ago, but I determined that was due to using my pinky on the ctrl key in emacs. That's a pretty well known problem, but I had used emacs for quite a few years without it being an issue. At that time, though, I had started using org-mode in a big way and was using a lot of C- bindings, and hitting it that often turned out to be a problem after all. Swapping ctrl and caps lock (and using my ring finger for it) fixed it entirely.
Even if the studies say that the difference in speed is marginal, I'm convinced based on my subjective experience that Dvorak is much more comfortable. That's difficult to quantify, though, and conducting a high-quality study to conclusively demonstrate it is extraordinarily difficult.
Personal anecdote, but I know a few people who got RSI/CTS and had a chance to watch them type --- and what they all had in common was how "stiff" their typing posture was, and the strength at which they hit the keys. They weren't fast either, probably in the 50-70wpm range. One of them, before he got RSI, even commented on how bad my posture was and how it'd "make it easier to injure myself".
I find that relaxing, as counterintuitive as it may sound at first, actually helps with typing speed and comfort. To me it makes sense that if you tense up all your muscles and punch the keys really hard, you're much more likely to injure yourself, slow down, and tire more quickly. I think the vast majority of keyboards also have too high of an actuation force, which also adds to the strain of typing. If you look at videos of fast typists (150wpm+), they don't look like they're exerting much effort at all; many of them aren't using the "standard" key-to-finger mapping, and a lot of them "smear" their hands all over the keyboard. Here's one example (not me):
Back when I was still an intern, I noticed that i was getting wrist pain and was instantly worries about getting carpel tunnel later so I went out and got a Microsoft Ergonomic keyboard.
I've since stuck with them and one of the things that I love about them is how they force you to use more of your fingers because it restricts what keys each hand has access to. I found that the way I typed changed, the 't', 'g', 'r' and 'f' keys I could no longer reach over with my left hand and vice versa for the other hand. This meant I had to switch my how I typed which inevitably switched which fingers I used to hit some fingers and forced me to use more of my hand for the typing (the "smear" as you call it)
> they all had in common was how "stiff" their typing posture was, and the strength at which they hit the keys.
Yeah! Its what I want to be researched. I play piano and guitar, and to do it successfully I had spent a lot of time learning to not apply more force than needed. Somtimes it needs to change a posture. I learned piano being a kid under a supervision of a teacher, so probably I don't remember the most of my learning troubles. But I learned guitar at my late teens and I remember clearly how I was corrected by a teacher, then corrected again and again. I oftentimes needed a break because of pain in muscles.
And I wonder might it be that in the most cases RSI is the consequences of the wrong posture or too much force applied. If it so, I suppose, switching layouts would help because of breaking a muscle memory and learning it anew, not because the new layout is better. The very RSI could influence learning: pain is a good teacher, it forces student to seek for painless ways to do things.
I am a pianist and have spent years trying to eliminate tension. It's a lifelong effort! But so worth it.
I recently saw a video of a concerto by cellist Bruno Philippe and was astonished to see how loosely he appeared to be holding the bow. It almost appeared that it would slip at any moment. Yet he was in total control and his tone was perfect!
I'm just gonna say that as someone typing 120+wpm for years now, I hate to see high actuation force vilified like this. It's my preference and I quite enjoy the higher force needed. Fewer accidental presses, and you can even lightly drum on your keys without them activating, which is fun and relaxing.
I went from cherry mx blues to cherry mx greens to kailh box navies and I love them to pieces.
I find that the people who swear by very light linear switches rarely type as fast as me, take that as you will. I think they fell down a strange rabbit hole.
The observation comes from my personal experience that actuation force seems to be almost directly proportional to typing speed and how long it takes to get tired. Maybe those with heavier fingers and/or stronger muscles would prefer a heavier keyfeel.
My fastest typing has been done on a ~40g rubber dome that actuates quite close to the top, and then has a very springy "cushion" past that (contrary to a lot of "common wisdom", they don't have to go all the way down to actuate --- that's what the "tit" on the bottom of the dome is for, and the longer and softer it is, the more cushion you get). Definitely not linear, but the bump is very subtle.
As for my speed: I can sustain around 150-160 for long periods, with brief (~10 second) bursts over 200.
I switched to Colemak 4 years ago or so, and I think it is one of the best investments of my time ever made. Yes, you will feel similar to that feeling you get when you write with your left hand (something like, why tf I'm doing this, if i'm perfectly capable of living my life as I have been doing before), but it is a skill that will benefit you all your life and I would recommend you to do so.
Still, it is not perfect. The pinky getting tired is something that happens, but it is still so much better than QWERTY. I don't think I will try that QWYRFM layout, as I haven't felt big pain since switching (before switching I felt some pain, but after, I slightly feel something, but after a lot of effort, and not enough to feel worried). And I think if I start feeling something to worry, there are some things I could do before, like using a split keyboard, or optimizing the desk height, before switching to another layout.
Again, try Colemak (or if you want to try any other layout, welcome). It will do good for you.
In my modest experience, most people who have issues with RSI and especially carpal problems, they almost universally use Emacs! It is such obvious thing.
I know you can remap keys or whatnot, but if you look at people experimenting with layouts and trying to resolve RSI problems, it is always Emacs.
I used emacs briefly, not long enough to say, but it seems pretty obvious where the problem is. I understand why people don't want to change their favorite editor.
The space cadet keyboard is often cited as inspiration for the Emacs combos. The Wikipedia article suggests its layout had mostly-symmetrical meta keys: https://en.wikipedia.org/wiki/Space-cadet_keyboard
I no longer use emacs, but that's the first thing I change on any new workstation. I hate that windows doesn't support an out of the box config for this.
The bigger problem is not the layout it’s the keyboard itself, if you have rsi switching to an ergodox can improve the situation, if this is not enough something like a dactyl manuform or a kinesis advantage 360 should do the trick.
This combined with light switches like tactile (MX brown) or linear (red) will reduce your pain immensely.
I wonder if anyone has studied the effects of different weights of switches. Personally I've always liked the lighter ones. The heavier ones felt uncomfortable and tiring. But I see people in the mechanical keyboard community praise the heavier switches, so I've always assumed that my hands are just weak.
The worst ones I used were ones that I got because the MK community doesn't consider browns to be able to be in the same class as other tactile switches because the bump is so weak. I ended up getting Zealio V2s, but I had to take them off my board after only a few days of use. The tactile bump is stupidly strong, but once you pass it there's almost no force pushing back. You're basically guaranteed on each press to bottom out with a ton of force. It was jarring and I was very afraid of what kind of damage that would do to my hands long term.
Well I like heavier switches because they allow me to fully rest my (rather large and strong) fingers on the keys. I'm actually typing this on 78g Zealio V2s, my favorite switch. I definitely don't bottom out the keys with a ton of force, and maybe like half the time don't bottom them out at all. By comparison, I bottom out on the iPad Magic Keyboard every time and with much more force, I would guess because there isn't as much tension pushing back on my fingers and there is less distance between the actuation and the bottom out to slow my fingers down.
I've had problems with the weight of my fingers actuating the switches when I tried some light linear switches. But for me the Cherry MX Browns have just enough resistance from the bump to avoid that issue without being tiring to my fingers.
> I wonder if anyone has studied the effects of different weights of switches. Personally I've always liked the lighter ones. The heavier ones felt uncomfortable and tiring. But I see people in the mechanical keyboard community praise the heavier switches, so I've always assumed that my hands are just weak
I don't know about studies, but I prefer membrane keyboards for this reason: far less strength is required.
Since changing to a mechanical keyboard, I've realized the opposite. A traditional membrane keyboard seems like much more effort. (I measured it at some point, by balancing coins on a key until the key activated, but I don't have enough coins to hand to repeat this.)
Laptop-style keys can be lighter -- I think too light -- but there are mechanical keys with similar weighting.
Next time you're in an electronics shop, see if there's a mechanical gaming keyboard on display with "Red" switches. (Note you probably don't need to press the key all they way to the bottom to activate it.) Switches lighter than this are available, but probably wouldn't be on display on a keyboard in a normal shop.
I love my mechanical keyboards, but membrane boards might still feel better to some even if they do require more force than some mechanical switch options. One specific benefit I can think of is that membrane boards naturally cushion your bottoming out and don't feel as mushy as O-rings or silenced switches can be.
I’m in the same box, I’ve tried Box Jade, Holy Pandas, MX Brown, Gateron Ink Black, Topre, MX Blues, Tangerine and Gateron Ink red. I also add some test switch zealio, etc…
But my favorite switch and daily driver are the gateron ink reds because they are so light.
It is absolutely the layout, but ergonomic keyboards are still great. I built a Pinky4 split keyboard and it didn't solve pain in my right hand caused by Dvorak. After learning Workman (50/50 balanced hand usage unlike any other layout I know of), the pain is gone.
The split helps by moving stuff like modifiers, space, and backspace to thumb keys, it just wasn't "enough" on its own.
Any keyboard layout that leads to more ulnar deviation is simply a no-go for me. So Dovrak and Colemak are out, evidently (never tried them, but TFA says they use the pinky more, so if I believe that then I must not use them). Ideally the pinky should not be used at all for typing.
I've had to teach myself to not use one-handed modifier keystrokes that involve ulnar deviation. The key to this, for me, has been the sticky keys accessibility feature -- it works fantastically well for me. If you've never heard of sticky keys before then I highly recommend you look it up and try it. It's not that that I type every modifier combination keystroke using sticky keys, but that using sticky keys helped me unlearn bad habits.
Another key to this, for me, is to use vim and not emacs. Modal editing means not relying on modifier keys -- conversely, non-modal editing means being extremely dependent on modifier keys. IMO reaching for modifier keys is probably half the RSI picture, with resting wrists while typing being the other half.
I think if you rule out Dvorak and Colemak due to pinky use, you should also rule out Qwerty. The best option would be to buy a keyboard with multiple thumb keys, so you can press e.g. enter and backspace with your thumb.
(Not using one-handed combinations is the general advice on any typing course -- A and M should use the shift keys on the right and left hand, respectively, as should ^A and ^M.)
I have learned to minimize pinky use on QWERTY, but you're right about thumb keys: I could really use control, shift, and alt modifiers right below the space bar. And you're right about typing courses, but.. I never took one, and I think that's true of most people. Perhaps more education would suffice to avoid most typist RSI.
> In Pan & Schleifer (1996), subjects experienced more arm discomfort/pain/fatigue the more keys they pressed while doing a data entry task.[23] Finally, Feng et al. is a recent (2021) cross-sectional study that found that “prolonged computer use time and working without breaks were associated with presence of wrist/hand symptoms”.[24] But of course correlation is not causation, and there are many potential confounders here, e.g. maybe people who spend lots of time at the computer exercise little, and it is the lack of exercise, not computer use, that causes problems.
I generally agree with this. I've found that I have less wrist pain now that (1) I've gotten a better mouse and (2) I work out my wrists and forearm occasionally. I think the keyboard is a much smaller cause, especially since a lot of repetitive stress comes more from gaming and use of mice for me.
I am a 100% laptop keyboard/trackpad user for both work & play for 10 years now and I haven't had any RSI since I stopped using a mouse (and paid close attention to overall posture, including having wrists straight but relaxed in both X and Y axis).
To your first citation, purely anecdotal but - over the years when I click on articles by programmers with serious RSI - the ones who identify their text editor are more often die-hard Emacs users (like Stallman). I don't think continual chording is a biologically respectful use of our analog digits. :-)
My own experiments with this point to similar conclusions. QWERTY is not laid out that badly, really, and for me (with no RSI problems, yet) the benefits of using the default, especially being able to type on other people's computers, far outweighed any benefit from the alternate layout (Workman). However, the excursion was completely worth it for me - I switched in the first place because I never learned to touch type properly, and figured it would be easier and more fun to learn from scratch on a new layout. That was true, and soon I found I'd entirely forgotten QWERTY, so I re-learned that from scratch, touch-typing, and am about as fast and accurate as I was on the alternate layout. Blog post after all that: https://patricksanan.org/personal/adventures-in-keyboarding-...
The fourth and fifth-least frequent English letters have the first and second-best location on the right hand, and a rarely-used symbol (;) has another.
Of the most frequent letters (ETAOIN), only one (A) is directly under a fingertip at a resting position, and two of them require the longest stretches (T, N).
Dvorak, by comparison, puts the least frequent letters on the bottom row: ;QJKX BMWVZ (PYGF are less frequent than M and W). It has all of ETAOIN on the middle row.
I've been typing english and only english on Dvorak for 6 years. For my needs, I am very happy with it. There is one point I would like to make that the author misses (and it's not unique to Dvorak).
Learning an alternative layout can make you a better typist for the sole reason of requiring effort to learn it. When most of us type with QWERTY, it is effortless, however we also carry with us all of the bad typing habits we've accumulated over time - all of the stuff we did wrong on a QWERTY board out of habit that our typing classes tried to condition out of us but it still remains. When you try to learn a new layout, you're putting in more effort than ever before into typing with correct form. That alone is a big factor for why non-QWERTY typists perform better.
I had an RSI issue where it was uncomfortable to reach for the H key and I was convinced existing layouts didn’t take this into account. I also thought that learning new keys was overrated so I attempted to generate keyboard changes with the least number of key changes from qwerty and with significantly less side-to-side motion. The results were very promising and then after trying them out for a bit it turned out Minimak-8 felt better to type on. Big shoutout to Minimak even if it didn’t help my H key :)
What the author may be missing is that keyboards are much, much better now than they were in the 1990s, when RMS was having problems. The touch is much lighter, the amount of force needed to depress the keys is much less. Hitting control-alt-meta-cokebottle-Q for everything to make Emacs work on an early 90s keyboard was a recipe for pain.
I've been typing on computers for almost as long as RMS has, don't have those problems.
I guess I should mention that I have my own self-taught touch typing style. I only use eight fingers, except for using my left pinky finger for the shift key. Works for me.
It's amazing how there seems to be more talk about layouts than ways to type less in the first place.
What we really need is a study on RSI and productivity vs overall workflow.
We already know Vim guys can edit code fast, what we need to know is who does best on a real world project with more debugging, research, and documentation reading than coding, how much IDE autocomplete helps, whether the 10 clicks and 2 minutes it takes me to do something is healthier than the 3 lines and 30 seconds it takes a CLI user, etc.
This position is how I've always typed, but I still had RSI problems. I switched to Dvorak with great, personal, success. If you watch a video of someone typing on Dvorak vs Qwerty [1], it's clear that the experience is very very different with any of these reduced motion layouts.
That video shows poor technique, it's the arms/hands that should hover over the keyboard, in one motion.
Stretching the fingers for reaching a given key is bad, Dvorak masks this by minimizing travel, which is simply a workaround / optimization for the wrong thing that long-term will result in the same problems (simply delayed a few years).
I could use an intentionally bad keyboard layout and it wouldn't matter. Arms are designed to be moved.
I don't think "masking" is the right way to look at it. Regardless of exact position, all hand motion is drastically reduced, for that hand position. There are many others on YouTube that show better technique, but this was the first video in the list that showed a simultaneous top down view.
I use a fairly ergonomic low profile split keyboard, and have no issues with my wrists, but I have some mild RSI in my index finger, near the knuckle joint. I have been trying to figure out how to fix this for a while, and haven't really found a solution, other than to switch to an alternative layout which is not as hard on the index finger. Anyone have a similar issue?
Not at all! RSI is complex because it can involve a hundred muscles and tendons. People can have inflammation and pain even in wrists, shoulders and elbows due to keyboard use.
Reducing finger travel by using Dvorak can be very beneficial. Obviously it will do little for things like ulnar pronation.
Myself, I had already switched to Colemak when I developed RSI due to heavy typing at my work. What did it for me was the unnatural position of my hands, wrists bent upwards and outwards in order to fit onto the "tiny" laptop keyboard.
The solution was simply to switch to a run-of-the-mill ergonomic keyboard (ERGO K860) and surprise, surprise, my RSI lessened and eventually disappeared in 2 weeks or so.
So, for me it was definitely the hand posture, NOT the layout that caused the problems!
I've been typing on QWERTY for over 40 years now, and have been an extensive Emacs user for decades, and luckily never had an issue with RSI. So, I wonder to what extent genetics play a role? What is the percentage of (heavy) QWERTY users that run into RSI issues at some point in their lives?
In in your position as well, and I'm not thinking that perhaps the RSI issues are more common for me peoples who use "correct" hand placement. I find Emacs to be very difficult to use with the proper hand placement, so perhaps it teaches users a different way to type.
I rarely use the pinkys to type, and have my hands shifted compared to the standard position (my index fingers are on f and h).
What this means is that I'm not always using the same fingers to type the same letters, allowing me to distribute the stress on my fingers. Of course, I end up up moving my hands more, but I don't see that as a problem.
I don't look at the keyboard while typing. I have no problems with typing entire sentences without looking down, and I type reasonably fast.
I used to type using the Dvorak layout when I was in high school. I gradually lost the ability to type in QWERTY at all. When I took the GRE I learned it would have to be in QWERTY, so I just switched back. I used QWERTY ever since.
I also tried to convert to an alternative layout, but then realized that on phones, qwerty is pretty objectively a good thing (due to large travels that help greatly there with accuracy). So I dissed the idea.
At least for me, phone muscle memory and keyboard muscle memory are two completely different things.
I've fully switched to the Neo2 Layout about 10 years ago and can't touch type on QWERTZ at all. But on my phone, using QWERT, I know where the letters are without thinking.
I had been programming since I was 11 years old (now 30 years old professional programmer). I’ve spent innumearable hours at computer keyboard. I can type QWERTY and Russian layout blindly both and I constantly use both.
Never had any issues with my wrists or finger pain.
What am I doing wrong?
Just to be clear I am not trying to say that people who use DVORAK are wrong. It’s just that you tend to forget there’s a huge population of people who due to their regional issues can’t even use DVORAK and most of them don’t have that problem with the wrists.
Maybe the only reason you had those issues is because the only time you actually learnt how to properly type on the keyboard is because you properly spent time learning how to blind type comfortably in DVORAK?
You're only 30. Give it 10 years. Not being snarky, the damage is being done, you're just not at that point yet. 38, been programming since the same age and I'm starting to feel it. Change your habits now and start focusing on ergonomics.
Point taken, but 30 is very young to be having issues with ergonomic injuries unless you're genetically predisposed to carpel tunnel or tendonitis or something.
As another commenter posted, it's not the typing that gets you, it's other stuff.
Like for example: you switch office space for a day, at the same time you got anew phone and you keep resting it on your pinky. You have a bad setup that day and think that it's no problem. Those in combination could easily cause a doctors visit because you can't use that pinky.
Do what you can to keep your hands in good shape and avoid weird static positions. Avoid weird movements. Try to use your back when lifting/pushing things, even small things like you coffee cup, lift close to the body not far away. Getting RSI is such a sucky thing have to deal with.
You are not doing anything wrong, you've just dodged bullets properly. So keep that going for you and make sure you keep on doing it!
I think there are stress/psychological factors. I developed RSI while being in a very bad living & working situation.
I've switched to COLEMAK and I still have pain. Does that mean COLEMAK doesn't work? Maybe. Or, since my pain level has stayed manageable it could mean I'd be worse off without it. I've been doing other things too. It's hard to reason on a sample of 1.
One place where typing speed does impact other people is synchronous messaging, like Slack. If you're asking someone for help, you're requesting their attention. It's better if you can type with sufficient speed to give plenty of information/context as to what you're talking about.
So, scientists didn't confirmed qwerty or things like that directly cause RSI/CTx...
But lets keep discussing that ! Let's circle from qwerty to dvorak and back again, with some minor route twiks to more exotic layouts...
Pleas stop...
R. is the problem ! 'R' like in 'repeat'. Programmer or pianist or pro-gamer or violinist or baker or whatever !
It's self abuse - Stellman wrote big parts of Emacs, gcc, some OS, other things while using mail and lists. jzw wrote Netscape, screensavers and other things. Other examples ?
I typed Dvorak for 10 years before I started getting RSI again in my right ulna. It turns out that the 'S' key being hit by the pinky in Dvorak caused me a lot of pain.
I ended up switching to BEAKL, a pretty obscure layout which is specifically made to alleviate stress on the pinky. It moves the relatively common S key to the right hand pointer finger, for instance.
It turns out that the theory underpinning the dvorak layout assumes that homerow hits are strictly better than upper and lower row, but the pinky and ring finger are such weak fingers that it's often better to move the index or middle fingers up or down a row than hit a key on the home row with the pinky.