The kind of person you probably are if you are reading Less Wrong spends a remarkable fraction of the day typing at a computer. As such, even a small increase in typing speed and skill can save you huge amount of time and effort.
This is a highly dubious claim. I (occupations: software engineer, student (CS major)) spend a remarkable fraction of the day at a computer… but do I spend most of that typing? I do not. I’m doing more typing right now, writing this comment, than I do in a much larger period of doing actual work. Even if you only look at the time I spend actively coding (rather than reading documentation / literature, thinking about a problem, debugging, tinkering, etc.) that’s still not mostly typing.
Furthermore, citation needed on the claim that touch-typing, as opposed to the way I type now, will save a “huge amount of time and effort”.
It is just as easy or easier to learn a better layout (like Dvorak or Colemak), which also will give you a bigger boost to your typing speed and efficiency.
For desk work that is not typing, look into a gaming keyboard and mouse. My drafting co-workers have bound short macros to the extra keyboard keys for frequently used commands, I am weighing the benefits to use the same approach for frequently used equations during design calculations.
A reasonable suggestion, though I find that the time required to bind the macros, then remember them, then remember to use them, is too much effort for me. That, of course, is up to personal preference.
Also: do you know of a gaming keyboard that is a Mac keyboard (presence of appropriate keys and layout) and has clicky-keys (a la the Apple Extended / Matias ProTouch Edit: got the name wrong, it’s the Matias Tactile Pro)?
Das Keyboard now comes in a real Mac version (they used to just have exchangeable keycaps and it’s never been quite ideal). It’s clicky and nice. It doesn’t have lunatic gaming ‘features’ like enormous rubber WASD though.
Yep, I’ve seen the Das Keyboard. Layout is still somewhat incorrect (see my comment here), even on the “Mac” version, though the fact that it’s based on the Model M (or… so the imply?) is cool. Of course it also doesn’t seem to have any extra keys that might be bound to macros or whatnot, which was the motivation for the original comment.
In general, gaming peripherals for Macs are hard to come by; Razer has at least 3 keyboards that allegedly support Mac OS X; you might have to modify or get over keys (mis)labeled for windows users. In general, most gaming keyboards like to have a very tactile feel that allows one to be certain of keypresses. I’m not sure exactly what feature you are describing with ‘clicky-keys’; are you looking for a long stroke, or for keys that register a keypress before bottoming out, and have significantly less resistance for the bottom portion of the stroke?
I’m describing the feel that you get with “mechanical-switch” keys (google it for images/explanations). The Apple Extended Keyboard (and Apple Extended Keyboard II), and, more recently, the Matias Tactile Pro (not ProTouch, whoops) are two keyboards that use such technology. Most other keyboards do not.
The Razer “for Mac” keyboards seem to have obviously non-Mac layouts, which rearrangement won’t fix. That’s a deal-breaker for me. The correct layout and the feel of a good keyboard is more important to me than additional keys or whatnot. My question was mostly due to curiosity.
I’m not sure what ‘layout’ issues you are referring to, but I’m not a Mac user and I have trouble adjusting to any laptop keyboard. I’m guessing that the layout issue is that you need four bucky keys to the left of the space bar and two to the right, instead of 3/4? Or is the location of the arrows and lack of the right half also critical?
EDIT: Mechanical keyboards can be tactile, noisy, both, or neither, depending on the specific nature of the switch. Which are important to you?
EDIT 2: What feature does the Matias Tactile Pro 3 keyboard lack that a ‘gaming’ keyboard would have?
What’s a “bucky key”...? (Edit: I see you were referring to this probably. Interesting, hadn’t heard the term before) Anyway, there are several layout issues. I’m pretty picky when it comes to keyboard layout. Um, let’s see. Correct number and layout of modifier keys (Control Option Command Spacebar Command Option Control; anything else is incorrect); correct numpad layout (18 keys, not 17, with Enter and 0 being the only two large ones); presence of volume control and eject keys above the numpad; shape of Return key (one-row height, not two-row); a properly sized backslash key… that ought to mostly cover it, I think.
I’m talking about the feel of the keyboard when I say that I prefer mechanical key switches.
As for what a gaming keyboard adds, I’m not the one who brought it up; wadavis mentioned them in this comment, citing extra keys to which macros might be bound. I’ve never used such a keyboard myself, I was just curious whether such a one might exist which would also fulfill my other criteria for a keyboard.
If all you want is programmable macros, then something like the Nostromo or Orbweaver could serve you well; every button and each direction of the 8-way hat switch can be changed to any input or macro possible from a USB device, including mouse clicks, multimedia commands, launching programs, or changing its own settings.
If your typical workflow can be reduced to 15-230 macros that you can keep straight, it could replace the keyboard as a primary input device. If you would benefit from 15-20 macros that you need to take a hand away from the keyboard to execute, it could serve as a useful addition. If you just want a keystroke combination that executes a series of commands, that’s probably better done in software.
I had good results with CopyPaste, a program which creates multiple clipboards you can store things in on a semipermanent basis. ctrl-shift-v-2 could be the command to paste the second stored clipboard, for instance (depending on your settings—it was very configurable). That only works for things that allow the use of the edit menu—not palette hotkey selections—but it could be a help.
While we don’t spend all our time in front of a computer typing, it does seem to represent a non negligible portion of our days. Assuming an hour a day of typing on average for the rest of your life, the time you will spend typing is tens of thousands of hours.
I’m currently learning Dvorak and it looks like it’s going to take about 30-40 hours to be able to type properly. So the gains in efficiency don’t have to be very significant to pay off.
To check how efficient the time investment is I checked my typing speed. Like you, probably, I’m not a touch typer but I felt like I was typing pretty decently before, and measured at 40 wpm on both of the two websites that I tried, with no mistakes. I’ll check my speeds with Dvorak once I’m done with the lessons, and again after a few months of practice, to settle this debate hopefully, but just from having done the first ten or so lessons I can already see that Dvorak is going to be a major improvement, if not in speed, definitely in terms of comfort.
You make an interesting point about likely spending over 10k hours typing over the course of the rest of our respective lives, although I note that even if you are right, I’d have to invest 30-40 hours now in order to learn to touch-type, whereas the gains would be spread out over a longer period. That said, please post your results when you get them, I am definitely interested in hearing about it.
I do note that you conflate two distinct issues: whether touch-typing is worth learning, and whether Dvorak is a meaningful (or any) improvement over QWERTY. I am definitely far more suspicious of the latter claim than the former (see my link in the grandparent for a thorough debunking).
I do note that you conflate two distinct issues: whether touch-typing is worth learning, and whether Dvorak is a meaningful (or any) improvement over QWERTY. I am definitely far more suspicious of the latter claim than the former (see my link in the grandparent for a thorough debunking).
Even the studies cited by the author in your link show a speed advantage of around 5% for Dvorak over Qwerty. Considering my point of the 10k hours, the payoff is more than worthwhile, before even taking RSI into consideration.
On a side note, one of the reasons I decided to learn touch-typing was because I have some free time at the moment and was looking for something else to do than read blog posts all day, so I totally agree with you that investing 30-40 hours now might not be the best for everyone… TDT probably recommends it though.
After having finished the basic course in Dvorak and touch typing for a few weeks now, here is an update on my results: I spent a total of 30 hours learning to touch type, but even once I could touch type properly, I was still really slow, at about 20 wpm immediately after finishing the lessons, half of my original speed. Ten days later, after forcing myself to avoid the QWERTY layout which resulted in some inconvenience, in particular with keyboard shortcuts, I am now typing at about 30 wpm in Dvorak, which is still significantly slower than my previous, unconventional but obviously not so bad, typing in QWERTY.
The idea that I will probably spend tens of thousands of hours typing in my life still stands, though, and the touch typing is getting more and more natural each day, I’ll try to update my results again after several months to see if there is actually a significant increase in typing speed over the long run.
On a side note, comfort is definitely better when touch typing “properly” in Dvorak than when typing “improperly” but faster in QWERTY, however this may be related to the way I positioned my hands on the keyboard rather than to the initial keyboard layout.
All of the evidence I have seen suggests that touchtyping is worth learning.
do I spend most of that typing? I do not.
To what extent is that because you’re a slow typist? (Do you know your wpm?)
So very citation needed on this one.
Dvorak, Colemak, or the superior QGMLWY generally will not increase typing speed for touchtypers, as typing speed for most applications is limited by thinking speed. They will increase efficiency, and one can estimate the reduced effort for any particular corpus with an effort model like carpalx’s, and so alternate layouts are primarily useful for people who want to prevent or manage repetitive stress injuries.
All of the evidence I have seen suggests that touchtyping is worth learning.
Links? :)
(Or, if this evidence is anecdotal or otherwise not easily linkable — please do elaborate!)
do I spend most of that typing? I do not.
To what extent is that because you’re a slow typist? (Do you know your wpm?)
I don’t know my wpm, but your question baffles me. How would my typing speed affect the fact that at some given moment I need to read several pages of documentation, sketch out a UI layout, look through code, think, etc.?
Your yourself say in your very next paragraph that “typing speed for most applications is limited by thinking speed” (and I think that’s only an upper bound on the practical limitation).
I don’t know what an “effort model” is, but I take from your comment that if I am not concerned about RSIs, Dvorak etc. should not interest me. Confirm/deny? Also, even assuming I am concerned about RSIs, do I understand correctly that the RSI prevention/management advantages of the alternate layouts you mention are for touch typists specifically, not just anyone typing in any way?
(Or, if this evidence is anecdotal or otherwise not easily linkable — please do elaborate!)
Most of it is anecdotal. The way I learned to touchtype was participating in chatrooms when I was younger; if you took too long to write sentences, the conversation would pass you by. So I quickly learned to type more quickly than I could talk. A more efficient way to learn is a blank keyboard. Here is an expensive one, or you can buy stickers for your current keyboard for $2 on Amazon, which also lets you learn letters one by one.
How would my typing speed affect the fact that at some given moment I need to read several pages of documentation, sketch out a UI layout, look through code, think, etc.?
The sort of activities you engage in will depend to some degree on the costs of those activities. If you can’t type quickly, you’re unlikely to participate in chatrooms or irc channels. The amount of journaling I do, say, might depend on whether I write my journal with a pen or with a keyboard, because it takes me far less time to press a key than to form a letter. If it takes fifteen minutes to jot down my record of the day rather than an hour, that might be enough to move the habit from not worthwhile to worthwhile.
I don’t know what an “effort model” is, but I take from your comment that if I am not concerned about RSIs, Dvorak etc. should not interest me. Confirm/deny?
Confirm. For me personally, it wasn’t worth the investment to switch from QWERTY to QGMLWY because transferring capped my typing speed at 7 wpm at a week, and the adaptation period typically runs ~2 months, suggesting I would be mostly out of commission for much longer than I thought was reasonable.
An effort model is an estimate of how much energy it takes / strain it puts on your fingers to press the key. Some fingers are stronger than others, and “home row” keys are easier to press than keys that require movement. (I move my hands around the keyboard, and so my “home row” is actually on several keyboard rows simultaneously, and moves based on what sentence I’m about to write, so the actual effort model for me is much more complicated than something like carpalx’s.)
This is a highly dubious claim. I (occupations: software engineer, student (CS major)) spend a remarkable fraction of the day at a computer… but do I spend most of that typing? I do not. I’m doing more typing right now, writing this comment, than I do in a much larger period of doing actual work. Even if you only look at the time I spend actively coding (rather than reading documentation / literature, thinking about a problem, debugging, tinkering, etc.) that’s still not mostly typing.
Furthermore, citation needed on the claim that touch-typing, as opposed to the way I type now, will save a “huge amount of time and effort”.
So very citation needed on this one. (Counter-citation: http://www.utdallas.edu/~liebowit/keys1.html.)
For desk work that is not typing, look into a gaming keyboard and mouse. My drafting co-workers have bound short macros to the extra keyboard keys for frequently used commands, I am weighing the benefits to use the same approach for frequently used equations during design calculations.
A reasonable suggestion, though I find that the time required to bind the macros, then remember them, then remember to use them, is too much effort for me. That, of course, is up to personal preference.
Also: do you know of a gaming keyboard that is a Mac keyboard (presence of appropriate keys and layout) and has clicky-keys (a la the Apple Extended / Matias ProTouch Edit: got the name wrong, it’s the Matias Tactile Pro)?
Update: I have recently purchased the Unicomp Spacesaver M[1], and it is everything I ever wanted from a keyboard. I can’t recommend it highly enough!
[1] Basically, it’s an IBM Model M—with buckling-spring keys—but with a Mac layout and a USB connection.
Das Keyboard now comes in a real Mac version (they used to just have exchangeable keycaps and it’s never been quite ideal). It’s clicky and nice. It doesn’t have lunatic gaming ‘features’ like enormous rubber WASD though.
Yep, I’ve seen the Das Keyboard. Layout is still somewhat incorrect (see my comment here), even on the “Mac” version, though the fact that it’s based on the Model M (or… so the imply?) is cool. Of course it also doesn’t seem to have any extra keys that might be bound to macros or whatnot, which was the motivation for the original comment.
In general, gaming peripherals for Macs are hard to come by; Razer has at least 3 keyboards that allegedly support Mac OS X; you might have to modify or get over keys (mis)labeled for windows users. In general, most gaming keyboards like to have a very tactile feel that allows one to be certain of keypresses. I’m not sure exactly what feature you are describing with ‘clicky-keys’; are you looking for a long stroke, or for keys that register a keypress before bottoming out, and have significantly less resistance for the bottom portion of the stroke?
I’m describing the feel that you get with “mechanical-switch” keys (google it for images/explanations). The Apple Extended Keyboard (and Apple Extended Keyboard II), and, more recently, the Matias Tactile Pro (not ProTouch, whoops) are two keyboards that use such technology. Most other keyboards do not.
The Razer “for Mac” keyboards seem to have obviously non-Mac layouts, which rearrangement won’t fix. That’s a deal-breaker for me. The correct layout and the feel of a good keyboard is more important to me than additional keys or whatnot. My question was mostly due to curiosity.
I’m not sure what ‘layout’ issues you are referring to, but I’m not a Mac user and I have trouble adjusting to any laptop keyboard. I’m guessing that the layout issue is that you need four bucky keys to the left of the space bar and two to the right, instead of 3/4? Or is the location of the arrows and lack of the right half also critical?
EDIT: Mechanical keyboards can be tactile, noisy, both, or neither, depending on the specific nature of the switch. Which are important to you?
EDIT 2: What feature does the Matias Tactile Pro 3 keyboard lack that a ‘gaming’ keyboard would have?
What’s a “bucky key”...? (Edit: I see you were referring to this probably. Interesting, hadn’t heard the term before) Anyway, there are several layout issues. I’m pretty picky when it comes to keyboard layout. Um, let’s see. Correct number and layout of modifier keys (Control Option Command Spacebar Command Option Control; anything else is incorrect); correct numpad layout (18 keys, not 17, with Enter and 0 being the only two large ones); presence of volume control and eject keys above the numpad; shape of Return key (one-row height, not two-row); a properly sized backslash key… that ought to mostly cover it, I think.
I’m talking about the feel of the keyboard when I say that I prefer mechanical key switches.
As for what a gaming keyboard adds, I’m not the one who brought it up; wadavis mentioned them in this comment, citing extra keys to which macros might be bound. I’ve never used such a keyboard myself, I was just curious whether such a one might exist which would also fulfill my other criteria for a keyboard.
If all you want is programmable macros, then something like the Nostromo or Orbweaver could serve you well; every button and each direction of the 8-way hat switch can be changed to any input or macro possible from a USB device, including mouse clicks, multimedia commands, launching programs, or changing its own settings.
If your typical workflow can be reduced to 15-230 macros that you can keep straight, it could replace the keyboard as a primary input device. If you would benefit from 15-20 macros that you need to take a hand away from the keyboard to execute, it could serve as a useful addition. If you just want a keystroke combination that executes a series of commands, that’s probably better done in software.
Off the top of my head, no.
I had good results with CopyPaste, a program which creates multiple clipboards you can store things in on a semipermanent basis. ctrl-shift-v-2 could be the command to paste the second stored clipboard, for instance (depending on your settings—it was very configurable). That only works for things that allow the use of the edit menu—not palette hotkey selections—but it could be a help.
While we don’t spend all our time in front of a computer typing, it does seem to represent a non negligible portion of our days. Assuming an hour a day of typing on average for the rest of your life, the time you will spend typing is tens of thousands of hours.
I’m currently learning Dvorak and it looks like it’s going to take about 30-40 hours to be able to type properly. So the gains in efficiency don’t have to be very significant to pay off.
To check how efficient the time investment is I checked my typing speed. Like you, probably, I’m not a touch typer but I felt like I was typing pretty decently before, and measured at 40 wpm on both of the two websites that I tried, with no mistakes. I’ll check my speeds with Dvorak once I’m done with the lessons, and again after a few months of practice, to settle this debate hopefully, but just from having done the first ten or so lessons I can already see that Dvorak is going to be a major improvement, if not in speed, definitely in terms of comfort.
You make an interesting point about likely spending over 10k hours typing over the course of the rest of our respective lives, although I note that even if you are right, I’d have to invest 30-40 hours now in order to learn to touch-type, whereas the gains would be spread out over a longer period. That said, please post your results when you get them, I am definitely interested in hearing about it.
I do note that you conflate two distinct issues: whether touch-typing is worth learning, and whether Dvorak is a meaningful (or any) improvement over QWERTY. I am definitely far more suspicious of the latter claim than the former (see my link in the grandparent for a thorough debunking).
Even the studies cited by the author in your link show a speed advantage of around 5% for Dvorak over Qwerty. Considering my point of the 10k hours, the payoff is more than worthwhile, before even taking RSI into consideration.
On a side note, one of the reasons I decided to learn touch-typing was because I have some free time at the moment and was looking for something else to do than read blog posts all day, so I totally agree with you that investing 30-40 hours now might not be the best for everyone… TDT probably recommends it though.
After having finished the basic course in Dvorak and touch typing for a few weeks now, here is an update on my results: I spent a total of 30 hours learning to touch type, but even once I could touch type properly, I was still really slow, at about 20 wpm immediately after finishing the lessons, half of my original speed. Ten days later, after forcing myself to avoid the QWERTY layout which resulted in some inconvenience, in particular with keyboard shortcuts, I am now typing at about 30 wpm in Dvorak, which is still significantly slower than my previous, unconventional but obviously not so bad, typing in QWERTY.
The idea that I will probably spend tens of thousands of hours typing in my life still stands, though, and the touch typing is getting more and more natural each day, I’ll try to update my results again after several months to see if there is actually a significant increase in typing speed over the long run.
On a side note, comfort is definitely better when touch typing “properly” in Dvorak than when typing “improperly” but faster in QWERTY, however this may be related to the way I positioned my hands on the keyboard rather than to the initial keyboard layout.
All of the evidence I have seen suggests that touchtyping is worth learning.
To what extent is that because you’re a slow typist? (Do you know your wpm?)
Dvorak, Colemak, or the superior QGMLWY generally will not increase typing speed for touchtypers, as typing speed for most applications is limited by thinking speed. They will increase efficiency, and one can estimate the reduced effort for any particular corpus with an effort model like carpalx’s, and so alternate layouts are primarily useful for people who want to prevent or manage repetitive stress injuries.
Links? :)
(Or, if this evidence is anecdotal or otherwise not easily linkable — please do elaborate!)
I don’t know my wpm, but your question baffles me. How would my typing speed affect the fact that at some given moment I need to read several pages of documentation, sketch out a UI layout, look through code, think, etc.?
Your yourself say in your very next paragraph that “typing speed for most applications is limited by thinking speed” (and I think that’s only an upper bound on the practical limitation).
I don’t know what an “effort model” is, but I take from your comment that if I am not concerned about RSIs, Dvorak etc. should not interest me. Confirm/deny? Also, even assuming I am concerned about RSIs, do I understand correctly that the RSI prevention/management advantages of the alternate layouts you mention are for touch typists specifically, not just anyone typing in any way?
Most of it is anecdotal. The way I learned to touchtype was participating in chatrooms when I was younger; if you took too long to write sentences, the conversation would pass you by. So I quickly learned to type more quickly than I could talk. A more efficient way to learn is a blank keyboard. Here is an expensive one, or you can buy stickers for your current keyboard for $2 on Amazon, which also lets you learn letters one by one.
The sort of activities you engage in will depend to some degree on the costs of those activities. If you can’t type quickly, you’re unlikely to participate in chatrooms or irc channels. The amount of journaling I do, say, might depend on whether I write my journal with a pen or with a keyboard, because it takes me far less time to press a key than to form a letter. If it takes fifteen minutes to jot down my record of the day rather than an hour, that might be enough to move the habit from not worthwhile to worthwhile.
Confirm. For me personally, it wasn’t worth the investment to switch from QWERTY to QGMLWY because transferring capped my typing speed at 7 wpm at a week, and the adaptation period typically runs ~2 months, suggesting I would be mostly out of commission for much longer than I thought was reasonable.
An effort model is an estimate of how much energy it takes / strain it puts on your fingers to press the key. Some fingers are stronger than others, and “home row” keys are easier to press than keys that require movement. (I move my hands around the keyboard, and so my “home row” is actually on several keyboard rows simultaneously, and moves based on what sentence I’m about to write, so the actual effort model for me is much more complicated than something like carpalx’s.)