I would also like to tentatively suggest an optimized keyboard layout such as Dvorak or Colemak, since the inconvenience is minimal if you’re starting from scratch, and there seems to be anecdotal evidence that they improve comfort and lessen RSIs in the long run, but if fretting about what layout to use causes you to procrastinate for even one day on learning to type already then you should forget I said anything.
Getting people to learn to type will be, however :-D
HOW THE HELL DO 80% OF THE COMPUTER-MAINLINING GEEKS I KNOW NOT KNOW HOW TO TYPE. HOW DO THEY NOT KNOW HOW TO USE THEIR PRIMARY MODE OF HUMAN INTERACTION. Figuring that out will be a study in human cognitive biases, for sure.
Yeah, there’s a reason i didn’t mention Dvorak or whatever ;-) So as not to put another “thing to do first” in the way. I know in person nobody at all who actually uses Dvorak. I can’t think of any Dvorak users amongst online friends I haven’t seen typing. (Perhaps there are some and they’ve just never said anything.)
I use Dvorak. It’s no faster and no more accurate, but it does tire out your fingers a whole lot less, and just typing one sentence in Dvorak will enable you to see why. I switched to Dvorak after a bout of RSI, and the RSI never came back.
If you work someplace that allows you basic administrator privileges, or just has a friendly systems administrator, it isn’t very difficult to change the keyboard layout in Windows. It can be set on a software level, or you could just bring a Dvorak keyboard in to work.
Unfortunately, half the jobs I’ve had wouldn’t allow this, so it’s not a guaranteed solution. And the software switch is only useful if you have a cover you can throw over the existing keyboard, or can touch-type sufficiently well.
Still, don’t think being employed eliminates the Dvorak option. I looked in to it just recently to make sure that learning Dvorak wouldn’t give me too much of a headache at work :)
That’s actually something I’ve never seen pointed out about Dvorak—every comparison seems to be about the speed relative to QWERTY. (Oh, the Wikipedia article mentions it in the first paragraph.)
Colemak user here. It didn’t magically improve my typing speed as I hoped, top speed is 70 wpm and used to be the same with qwerty. I’m pretty sure it’s more ergonomic to type with than qwerty, and I do have some wrist problems, so I’m going to stick with it.
I don’t think non-mainstream layouts are something people should feel obliged to adopt unless they are having wrist problems. Beyond the ergonomics, it’s mostly a weird thing to learn for fun.
Didn’t like Dvorak because it makes you type ‘ls’ with your right pinky, and I type ‘ls’ a lot on unixlike command line shells.
It occurs to me that ‘l’ is also ‘move right’ in vim. I think I find my rightmost three fingers hovering on the top row when I move about for this reason. Wonder if I should try to remap those movement keys...
The vim movement keys actually work surprisingly well in Dvorak. Up/Down are next to each other on your left hand, right/left are on the appropriate sides of your right hand.
The nice thing about keyboard layouts, now that we have reprogrammable computers, is that there’s little need to have holy wars over them. Having more people use the same layout is mostly inconsequential to a single user of the layout. It’s very different for operating systems, programming languages and programs, where a lack of users means a lack of support and a slow slide into obscurity and eventual unusability.
Eliezer uses Dvorak, or at least used to four years ago:
I can personally testify that Dvorak seems to be much easier on the fingers than Qwerty—but this is not surprising, since if Dvorak really were inferior to Qwerty, it would soon cease to exist. (Yes, I am familiar with the controversy in this area—bear in mind that this is a politically charged topic since it has been used to make accusations of market failure. Nonetheless, my fingers now sweat less, my hands feel less tired, my carpal tunnel syndrome went away, and none of this is surprising because I can feel my fingers traveling shorter distances.)
Except I’ve been typing for a living for 13 years on QWERTY and never had carpal tunnel syndrome. It’s not clear to me that it has anything to do with keyboard layout.
Reasons one may not have carpal tunnel syndrome may be:
1) independent of their keyboard layout, e.g. their carpal tunnels are very resilient, or they may not type enough to injure them;
2) dependent on the keyboard layout, e.g. for the typing one does one layout may be “efficient” enough not to trigger the syndrome.
The observation that one never had CTS doesn’t separate the two hypotheses (i.e., you can’t tell if you never had carpal tunnel because of 1 or 2).
My personal experience, as well as reports from others (e.g. Eliezer), is that typing on QWERTY did cause CTS, and after switching to Dvorak (for many years now), without any other visible change in typing (quantity or kind) the symptoms disappeared.
From this evidence, the conclusion is quite clear that Dvorak is better for CTS than QWERTY. To be unclear about it you’d need to also have observations of people that had CTS with Dvorak but not with QWERTY. (However, it’s also clear that QWERTY is enough for some people, and that you’re likely in that category.)
(Of course, the conclusion is “clear”, as I said, based on the evidence cited. It’s not a lot of evidence, so it doesn’t mean that the conclusion is definite in general. I just pointed out that you have more evidence than your personal experience that you’re ignoring.)
(ETA: Also, it appears that you don’t quite need to worry about it. Similarly, I picked Dvorak when I had CTS, my CTS went away, and I don’t need to worry about layouts better than Dvorak. That doesn’t mean I’m not clear about Dvorak being less efficient than other layouts.)
To be unclear about it you’d need to also have observations of people that had CTS with Dvorak but not with QWERTY. (However, it’s also clear that QWERTY is enough for some people, and that you’re likely in that category.)
Incorrect. As QWERTY is the standard, most people who have no problem with QWERTY don’t switch.
Therefore, people for whom QWERTY is more efficient than Dvorak are highly unlikely to ever use Dvorak enough to develop problems (such as CTS). If, say, 10% of the population was better off with Dvorak and 90% was better off with QWERTY, you still wouldn’t expect to see people developing CTS with Dvorak, then going to QWERTY, because most people start with QWERTY.
I’m not saying that QWERTY is better for anyone than Dvorak (personally the only reason I stopped using Dvorak was because I couldn’t work out how to change the commands for ctrl-c, ctrl-x, ctrl-z, ctrl-s etc. to be in the same positions, rather than spread all over the keyboard) merely that it’s a perfectly reasonable possibility given the evidence presented.
It’s easy to learn. You can still retain qwerty proficiency. It does feel nicer for typing English. It doesn’t help programming. It’s annoying to use multiple/public computers.
There are quite a few layouts that may be better than Dvorak. But probably not by enough to justify the extra effort of choosing one.
I first learned how to touch type on Dvorak, but switched to qwerty when I went to college so I wouldn’t have issues using other computers. I found that I could not maintain proficiency with both layouts. One skill just clobbered the other.
Maybe that’s true once you try to get extremely fast with both.
Since elementary school typing class, I’ve been 80+ wpm qwerty.
I only learned and used dvorak up to about 50-60 wpm. Perhaps I never could have built maximum competence in both. I definitely noticed some mode-switching overhead.
But probably not by enough to justify the extra effort of choosing one.
I disagree. (And am biased as it gets.)
Qwerty is really pretty bad. But looking out for the available once might make a difference in experienced typeability.
The network effect of keyboards is marginally. Some are preinstalled in your favorite OS, some are not. But otherwise you end up with about the same effort for relearning and explaining to other computer users why yours works different, that you can really invest a few hours to first look up which layout suits you the most.
Dvorak is probably worth learning. I’m saying that (except if you spend most of your time typing non-English text, e.g. some programming language that has much more typing time than thinking time), it’s probably not worth finding a more optimal layout than Dvorak.
In fact, if you have examples of the types of text you most often type, you can find a nonstandard layout using computer optimization, which is what I was thinking of.
My rough view is that for typical English text, the efficiencies are:
qwerty: 85%
dvorak: 97%
(all other layouts): at most 100%
So it’s better to just learn Dvorak now than to choose something that has more implementation effort than choosing the layout from a menu (iphone, windows, mac, and unix will probably all have a menu that includes a dvorak mode, but not some more esoteric 99% efficient layout for your workload).
The efficiency numbers I give are in terms of actual trained speed and accuracy. In fact, by metrics like “finger miles”, Dvorak is dramatically more efficient than qwerty. That’s not what I optimize, though. I am skeptical that RSI risk scales mainly with “daily finger miles” (I had pretty severe wrist/shoulder RSI for over a year in the past).
This has made me wonder why I don’t use Dvorak myself. I think it’s mostly because I didn’t learn the full keyset I use for programming. And I would probably prefer one of the more-like-qwerty punctuation layouts (that mainly rearranges letter keys), but I don’t want to decide …
I’m guilty of not really following my own advice, so I welcome a refutation :)
I try my best, but I also am very biased and probably not an expert.
The main effort in switching layouts is a fixed re:training time, and then some minor hassles when interfacing with other people. It is not the time to install the new driver!
(I usually regroup the keys on my keyboards, but thats also a fixed time per keyboard bought and some low level fun.)
There is no research I am aware of that confirms the RSI/finger movement connection. All pleasure derived from my Layout of choice is purely subjective.
The point I tried to make is that it pays to choose well before deciding to spend the effort for training a new layout. Take a few hours to reach a somewhat reasonable decision and then go about it, instead of just following a subcultural trend.
I think you underestimate the possible benefits gained from a better designed keyboard. There is a lot of space at the top.
What makes me like Neo tremendously besides the optimization are the additional layers. Pay attention to the 3rd layer in the overview. All brackets are nicely available in the center field.I would like to see that tacked onto any other layout one might choose.
(I am also disappointed that professional researchers into work ergonomy did not attack this topic on their own. The layouts I checked out seem to be fan projects. And with current technology it is almost trivial to calculate a good one at home.)
And being somewhat of a language buff I can type all latin-based languages natively from my keyboard. Without installing anything extra or switching layouts multiple times. That might not be too important for all-english writers, but for me that means some benefit with my 3rd language and possible later ones.
I would expect that someone interested in the topic is not immediately aware of the possibilities offered.
Actually before relearning I calculated the expected benefits and came out with what economists call a black 0 meaning some minor benefits. It is more valuable to retrain to another layout when young, and when you expect to write a lot of text over your lifetime. Programmers might be on the edge of not benefiting too much from it. But I find the experience of having to relearn highly interesting in itself. It helps in empathizing with computer newbies.
If you decide to go for a new layout, give at least Colemak a good thorough look. If not, no harm done.
Disclaimer: I use NEO since 2006, after a brief try with dvorak]
I agree that subjective pleasure with your choices is very important.
I just remembered another reason I chose to stay qwerty—emacs keybindings and video games (although I’m video game abstinent for the past few months). The default letters-as-commands mappings would have to be changed or positionally relearned for each such application I’m familiar with (similar: ctrl-z x c v in windows). Overall I didn’t feel like investing the effort to resolve the annoyance, but I guess I wish now that I had made the investment; I’d probably enjoy the result as you do.
The effort of installing a new layout isn’t much, you’re right (unless you hop computers often). It just might be if you use especially limited devices (does the iphone/ipad keyboard even support arbitrary layouts?) that you sometimes need to qwerty anyway.
The NEO developergroup payed attention to many of those. They also collected common sets of two or three letters from common applications. So the Smiley becomes just one roll over three buttons. I probably reap some benefits from that once I get back into Lateχ. In general I like to use tools that are optimized over my current horizon and can surprise me with thinks already put in way after I started to use them.
I guess I wish now that I had made the investment
You can make the investment at any time you choose. Once you did the calculation changes (thanks to sunken cost) but before that its a matter of finding a convenient time space. Like when one is sick at home, or in holidays.
Iphone/ipad does not have Colemak or generic support for different layouts. Not sure about Dvorak.
I know at least 2 Dvorak users, 1 Colemak user, and 1 NEO user personally, and a few who are interested to learn.
For anyone interested in switching layouts: skip Dvorak and go to one of the newer computer optimized layouts right away. I found it an interesting experience to have to re:learn how to type.
The author uses dynamic programming to calculate the various costs involved with typing (like finger movement, distance from home row, etc) and uses that to generate better layouts via simulated annealing. I thought it was a nicely quantitative take on a subject that is usually so subjective.
Although he had the right idea, I think this author’s analysis was rather poor. I don’t think he did a good job of modelling the importance of different kinds of typing strains. I like Colemak a lot better.
Upvoting this did not seem adequate.
I would also like to tentatively suggest an optimized keyboard layout such as Dvorak or Colemak, since the inconvenience is minimal if you’re starting from scratch, and there seems to be anecdotal evidence that they improve comfort and lessen RSIs in the long run, but if fretting about what layout to use causes you to procrastinate for even one day on learning to type already then you should forget I said anything.
Getting people to learn to type will be, however :-D
HOW THE HELL DO 80% OF THE COMPUTER-MAINLINING GEEKS I KNOW NOT KNOW HOW TO TYPE. HOW DO THEY NOT KNOW HOW TO USE THEIR PRIMARY MODE OF HUMAN INTERACTION. Figuring that out will be a study in human cognitive biases, for sure.
Yeah, there’s a reason i didn’t mention Dvorak or whatever ;-) So as not to put another “thing to do first” in the way. I know in person nobody at all who actually uses Dvorak. I can’t think of any Dvorak users amongst online friends I haven’t seen typing. (Perhaps there are some and they’ve just never said anything.)
I use Dvorak. It’s no faster and no more accurate, but it does tire out your fingers a whole lot less, and just typing one sentence in Dvorak will enable you to see why. I switched to Dvorak after a bout of RSI, and the RSI never came back.
del
If you work someplace that allows you basic administrator privileges, or just has a friendly systems administrator, it isn’t very difficult to change the keyboard layout in Windows. It can be set on a software level, or you could just bring a Dvorak keyboard in to work.
Unfortunately, half the jobs I’ve had wouldn’t allow this, so it’s not a guaranteed solution. And the software switch is only useful if you have a cover you can throw over the existing keyboard, or can touch-type sufficiently well.
Still, don’t think being employed eliminates the Dvorak option. I looked in to it just recently to make sure that learning Dvorak wouldn’t give me too much of a headache at work :)
del
That’s actually something I’ve never seen pointed out about Dvorak—every comparison seems to be about the speed relative to QWERTY. (Oh, the Wikipedia article mentions it in the first paragraph.)
Colemak user here. It didn’t magically improve my typing speed as I hoped, top speed is 70 wpm and used to be the same with qwerty. I’m pretty sure it’s more ergonomic to type with than qwerty, and I do have some wrist problems, so I’m going to stick with it.
I don’t think non-mainstream layouts are something people should feel obliged to adopt unless they are having wrist problems. Beyond the ergonomics, it’s mostly a weird thing to learn for fun.
Didn’t like Dvorak because it makes you type ‘ls’ with your right pinky, and I type ‘ls’ a lot on unixlike command line shells.
It occurs to me that ‘l’ is also ‘move right’ in vim. I think I find my rightmost three fingers hovering on the top row when I move about for this reason. Wonder if I should try to remap those movement keys...
The vim movement keys actually work surprisingly well in Dvorak. Up/Down are next to each other on your left hand, right/left are on the appropriate sides of your right hand.
that never occurred to me. I may write some bash aliases with a view to reducing long movements today.
The Wikipedia article on keyboard layouts is very interesting and informative.
The nice thing about keyboard layouts, now that we have reprogrammable computers, is that there’s little need to have holy wars over them. Having more people use the same layout is mostly inconsequential to a single user of the layout. It’s very different for operating systems, programming languages and programs, where a lack of users means a lack of support and a slow slide into obscurity and eventual unusability.
Eliezer uses Dvorak, or at least used to four years ago:
Except I’ve been typing for a living for 13 years on QWERTY and never had carpal tunnel syndrome. It’s not clear to me that it has anything to do with keyboard layout.
Reasons one may not have carpal tunnel syndrome may be: 1) independent of their keyboard layout, e.g. their carpal tunnels are very resilient, or they may not type enough to injure them; 2) dependent on the keyboard layout, e.g. for the typing one does one layout may be “efficient” enough not to trigger the syndrome.
The observation that one never had CTS doesn’t separate the two hypotheses (i.e., you can’t tell if you never had carpal tunnel because of 1 or 2).
My personal experience, as well as reports from others (e.g. Eliezer), is that typing on QWERTY did cause CTS, and after switching to Dvorak (for many years now), without any other visible change in typing (quantity or kind) the symptoms disappeared.
From this evidence, the conclusion is quite clear that Dvorak is better for CTS than QWERTY. To be unclear about it you’d need to also have observations of people that had CTS with Dvorak but not with QWERTY. (However, it’s also clear that QWERTY is enough for some people, and that you’re likely in that category.)
(Of course, the conclusion is “clear”, as I said, based on the evidence cited. It’s not a lot of evidence, so it doesn’t mean that the conclusion is definite in general. I just pointed out that you have more evidence than your personal experience that you’re ignoring.)
(ETA: Also, it appears that you don’t quite need to worry about it. Similarly, I picked Dvorak when I had CTS, my CTS went away, and I don’t need to worry about layouts better than Dvorak. That doesn’t mean I’m not clear about Dvorak being less efficient than other layouts.)
Incorrect. As QWERTY is the standard, most people who have no problem with QWERTY don’t switch.
Therefore, people for whom QWERTY is more efficient than Dvorak are highly unlikely to ever use Dvorak enough to develop problems (such as CTS). If, say, 10% of the population was better off with Dvorak and 90% was better off with QWERTY, you still wouldn’t expect to see people developing CTS with Dvorak, then going to QWERTY, because most people start with QWERTY.
I’m not saying that QWERTY is better for anyone than Dvorak (personally the only reason I stopped using Dvorak was because I couldn’t work out how to change the commands for ctrl-c, ctrl-x, ctrl-z, ctrl-s etc. to be in the same positions, rather than spread all over the keyboard) merely that it’s a perfectly reasonable possibility given the evidence presented.
My brother has used Dvorak for the past 10 years.
It’s easy to learn. You can still retain qwerty proficiency. It does feel nicer for typing English. It doesn’t help programming. It’s annoying to use multiple/public computers.
There are quite a few layouts that may be better than Dvorak. But probably not by enough to justify the extra effort of choosing one.
I first learned how to touch type on Dvorak, but switched to qwerty when I went to college so I wouldn’t have issues using other computers. I found that I could not maintain proficiency with both layouts. One skill just clobbered the other.
Maybe that’s true once you try to get extremely fast with both.
Since elementary school typing class, I’ve been 80+ wpm qwerty.
I only learned and used dvorak up to about 50-60 wpm. Perhaps I never could have built maximum competence in both. I definitely noticed some mode-switching overhead.
I disagree. (And am biased as it gets.) Qwerty is really pretty bad. But looking out for the available once might make a difference in experienced typeability.
The network effect of keyboards is marginally. Some are preinstalled in your favorite OS, some are not. But otherwise you end up with about the same effort for relearning and explaining to other computer users why yours works different, that you can really invest a few hours to first look up which layout suits you the most.
You may have misunderstood me.
Dvorak is probably worth learning. I’m saying that (except if you spend most of your time typing non-English text, e.g. some programming language that has much more typing time than thinking time), it’s probably not worth finding a more optimal layout than Dvorak.
In fact, if you have examples of the types of text you most often type, you can find a nonstandard layout using computer optimization, which is what I was thinking of.
My rough view is that for typical English text, the efficiencies are:
qwerty: 85%
dvorak: 97%
(all other layouts): at most 100%
So it’s better to just learn Dvorak now than to choose something that has more implementation effort than choosing the layout from a menu (iphone, windows, mac, and unix will probably all have a menu that includes a dvorak mode, but not some more esoteric 99% efficient layout for your workload).
The efficiency numbers I give are in terms of actual trained speed and accuracy. In fact, by metrics like “finger miles”, Dvorak is dramatically more efficient than qwerty. That’s not what I optimize, though. I am skeptical that RSI risk scales mainly with “daily finger miles” (I had pretty severe wrist/shoulder RSI for over a year in the past).
This has made me wonder why I don’t use Dvorak myself. I think it’s mostly because I didn’t learn the full keyset I use for programming. And I would probably prefer one of the more-like-qwerty punctuation layouts (that mainly rearranges letter keys), but I don’t want to decide …
I’m guilty of not really following my own advice, so I welcome a refutation :)
I try my best, but I also am very biased and probably not an expert.
The main effort in switching layouts is a fixed re:training time, and then some minor hassles when interfacing with other people. It is not the time to install the new driver! (I usually regroup the keys on my keyboards, but thats also a fixed time per keyboard bought and some low level fun.)
There is no research I am aware of that confirms the RSI/finger movement connection. All pleasure derived from my Layout of choice is purely subjective.
The point I tried to make is that it pays to choose well before deciding to spend the effort for training a new layout. Take a few hours to reach a somewhat reasonable decision and then go about it, instead of just following a subcultural trend.
I think you underestimate the possible benefits gained from a better designed keyboard. There is a lot of space at the top.
What makes me like Neo tremendously besides the optimization are the additional layers. Pay attention to the 3rd layer in the overview. All brackets are nicely available in the center field.I would like to see that tacked onto any other layout one might choose.
(I am also disappointed that professional researchers into work ergonomy did not attack this topic on their own. The layouts I checked out seem to be fan projects. And with current technology it is almost trivial to calculate a good one at home.)
And being somewhat of a language buff I can type all latin-based languages natively from my keyboard. Without installing anything extra or switching layouts multiple times. That might not be too important for all-english writers, but for me that means some benefit with my 3rd language and possible later ones.
I would expect that someone interested in the topic is not immediately aware of the possibilities offered.
Actually before relearning I calculated the expected benefits and came out with what economists call a black 0 meaning some minor benefits. It is more valuable to retrain to another layout when young, and when you expect to write a lot of text over your lifetime. Programmers might be on the edge of not benefiting too much from it. But I find the experience of having to relearn highly interesting in itself. It helps in empathizing with computer newbies.
If you decide to go for a new layout, give at least Colemak a good thorough look. If not, no harm done.
Disclaimer: I use NEO since 2006, after a brief try with dvorak]
I agree that subjective pleasure with your choices is very important.
I just remembered another reason I chose to stay qwerty—emacs keybindings and video games (although I’m video game abstinent for the past few months). The default letters-as-commands mappings would have to be changed or positionally relearned for each such application I’m familiar with (similar: ctrl-z x c v in windows). Overall I didn’t feel like investing the effort to resolve the annoyance, but I guess I wish now that I had made the investment; I’d probably enjoy the result as you do.
The effort of installing a new layout isn’t much, you’re right (unless you hop computers often). It just might be if you use especially limited devices (does the iphone/ipad keyboard even support arbitrary layouts?) that you sometimes need to qwerty anyway.
The NEO developergroup payed attention to many of those. They also collected common sets of two or three letters from common applications. So the Smiley becomes just one roll over three buttons. I probably reap some benefits from that once I get back into Lateχ. In general I like to use tools that are optimized over my current horizon and can surprise me with thinks already put in way after I started to use them.
You can make the investment at any time you choose. Once you did the calculation changes (thanks to sunken cost) but before that its a matter of finding a convenient time space. Like when one is sick at home, or in holidays.
Iphone/ipad does not have Colemak or generic support for different layouts. Not sure about Dvorak.
I know at least 2 Dvorak users, 1 Colemak user, and 1 NEO user personally, and a few who are interested to learn.
For anyone interested in switching layouts: skip Dvorak and go to one of the newer computer optimized layouts right away. I found it an interesting experience to have to re:learn how to type.
There’s a really interesting comparison of popular keyboard layouts and proposed optimizations here: http://mkweb.bcgsc.ca/carpalx/
The author uses dynamic programming to calculate the various costs involved with typing (like finger movement, distance from home row, etc) and uses that to generate better layouts via simulated annealing. I thought it was a nicely quantitative take on a subject that is usually so subjective.
Although he had the right idea, I think this author’s analysis was rather poor. I don’t think he did a good job of modelling the importance of different kinds of typing strains. I like Colemak a lot better.