Hacker News Reader

Comment by donatj | Hacker News Reader

"[T]he evidence in the standard history of Qwerty versus Dvorak is flawed and incomplete. [..] The most dramatic claims are traceable to Dvorak himself; and the best-documented experiments, as well as recent ergonomic studies, suggest little or no advantage for the Dvorak keyboard."

It's an urban myth that Dvorak is any better than qwerty.

View on HN

I used to be around 80 wpm with qwerty, now I'm 65 on dvorak, but it's much more comfortable.

It's hard to explain, but it feels satisfying to type. When I think about a word I'll sometimes think about what motions it takes to type it (on dvorak) and it just feels comfortable in my head. No regrets about the loss in typing speed.

The worst part about using dvorak is that no software is (or ever will be) designed for dvorak. Vim and emacs still work great, most shortcuts are phonetic and jk are where cv are, so I don't really mind.

I have been waiting for someone to sit down and redesign the keyboard layout for programming, and maybe this is it.

I've been typing on Programmer Dvorak[0] for 3 over years.

[0] https://www.kaufmann.no/roland/dvorak/ it comes with most (all?) Linux systems `setxkbmap -layout us -variant dvp`

I've been on and off with Dvorak since about 16 years.

The feeling is consistent, when I do a lot of typing during the day with QWERTY my hands ache, with Dvorak they don't. So as you say, as long as the typing speed is ok, there's a huge win in how comfortable it is. For me the best way of separating between the two is how the motions are as well, with QWERTY everything is clustered and with Dvorak each key is on each hand, so much more of a hand dance, which in itself is quite nice.

I do remember the first time I introduced myself to it though. Removed all the caps on the keyboard and placed them accordingly, found myself unable to find the correct letters even though I searched the keyboard. Turns out every letter was under the hand that still rested on the keyboard ;) Took ages just to write common sentences, while the brain went crazy. Worth it now though!

The software support is better nowadays, used to be horrible with VNC and the likes. With NoVNC catching wind all those problems are gone though, even on things like IPMI with HTML5 consoles! RDP still blows my mind by changing the keyboard layout to whatever is running on the target machine.

Same, with Colemak. I actually type a little faster now, but the major benefit is how comfortable it is. Hardly have to leave the home row.

+1 for colemak.

I've used all of Qwerty, Dvorak, Programer Dvorak, and Colemak. I think Colemak is the clear winner in terms of comfort. I also don't care much about speed, I just want to not have my hands ache by the end of the day.

I experience the same satisfaction when touch typing and thinking about typing motions, but I've always used QWERTY. Do you have any idea why you didn't experience this until you moved to Dvorak?

I think one problem I encountered prior to switching to Dvorak was that QWERTY encouraged me to make bad touch typing habits like holding my right hand further toward the top row rather than at the home. This exacerbated my prior inability to pick up ambidextrous use of shift and ctrl.

My error rate is now tiny, backspace is for bad writing rather than bad typing. It is really comfortable to type on Dvorak in all manner of postures. I can type with my hands coming at the keyboard sideways, way above, way below, or while dancing or walking with the keyboard on a moving dolly; I've had no strain issues on any keyboard from doing this sort of thing regularly, or from sustained typing while sitting in crappy office chairs behind desks at the wrong height.

I think it might have something to do with the complete effectiveness of the home row. Because the home row reigns supreme in Dvorak, changes in posture aren't compounded by changes is hand positioning on the keyboard to hit certain motions.

My experience is similar to yours, except I am not looking for anything else. I switched to Dvorak 8 years ago and I haven't looked back since. My speed was lower and might still be but I don't care, typing is a delight and relative to the amount of time I spend on thinking and reasoning I think the overhead of any loss in typing speed is probably insignificant. Vim is the only keyboard control that matters to me anyway. (Though Vim is far from the only program I use.)

I feel it's important to mention that much of my love for Dvorak comes from the grid layout keyboard I've been using for all these years (wore one out though but bought a couple more to replace it), and which remains my primary keyboard both at home and at work.

I've experimented with grid layout; I like the Kinesis Advantsge. Would you mind sharing? Which brand works well for you?

TypeMatrix 2030 USB

Not OP but I have a TypeMatrix that I really like.

Try a Kinesis Advantage: https://jakeseliger.com/2009/07/20/kinesis-advantage. I think it's likely to be a bigger improvement than QWERTY vs Dvorak.

I have one and like it a lot, but it doesn't work in my bag or at the coffee shop.

Awesome. Let me know what you think after you play with it. I'm totally open to making changes.

>It's an urban myth that Dvorak is any better than qwerty.

I think you're wrong. The Carpal X keyboard layout optimizer consistently ranks Dvorak as better than QWERTY for "effort" (which takes into account things like finger travel distance and how good we are at using certain fingers).

You can see the findings here: http://mkweb.bcgsc.ca/carpalx/?typing_effort

Their conclusion was to determine which layout is the best - their own layout, which I use, called QGMLWB was determined to be the one which requires the least typing effort. I'd recommend people try it.

A quote supporting the advantages of Dvorak:

>First, Dvorak is very good at keeping fingers on home row - 71% of keystrokes land there (compare this with 34% for QWERTY). This alone is worth the price of admission. Dvorak bottom row usage is very low at 9% (15% for QWERTY). Dvorak favours the right hand by 14% (QWERTY favours left by 15%). Dvorak has more uniform finger usage and makes greater use of the pinky (18% vs QWERTY's 10%).

http://mkweb.bcgsc.ca/carpalx/?dvorak

> Dvorak favours the right hand by 14% (QWERTY favours left by 15%).

Does that mean Dvorak is better for right handed people? Any left handers with experience here?

Greater use of pinky is a minus. It's your weakest finger.

For Dvorak, aren’t the frequently used keys (for English) under the fingers on the home row? You can measure travel distance of the fingers, for example, on a novel, etc

http://patorjk.com/keyboard-layout-analyzer/#/main

Xah Lee has a heat map:

http://xahlee.info/kbd/keyboard_dvorak_layout.html

I think the claim is that whatever limits typing speed is not key travel distance, but rather maybe coordination in your brain or some other factor. Anecdotally I would get tired typing qwerty way faster than I do typing dvorak, though some of this is just the fact that my form is much better in dvorak.

Yeah hard to account for that confounding factor. The alt layout is most people's second, learned much more deliberately.

The heat map is interesting in that it shows clearly the reason I moved from Dvorak to Colemak after 5 years of the former: Dvorak has a distinct bias toward the right hand, and particularly the right pinky.

How is this relevant to the article?

And, you rarely hear of Dvorak typists that will say they switched for speed and stayed for the speed. It’s almost always that they switch for the myth of speed and claims of comfort, but stayed for the comfort.

I thought the urban myth was about Dvorak being faster than qwerty.

For english language Dvorak does increase center row usage vs. top and bottom row. And for programming there is programmers dvorak.

The absolute worst keyboard layout i've seen for programming is the german layout. Using modeswitch/altgr(yeah i know it's not quite the same) is absolutely terrible.

I'm not sure if Dvorak is really faster, but I do think it keeps your fingers on the home row more. At this point the main reason I prefer Dvorak is that it lets me have more modifier keys on the left side of the keyboard (without moving to the number row).

I don't think layout matters for speed unless we're talking top 1% of typists. Go on YouTube. Tons of people typing 100+ wpm with two fingers.

I type 60-70 WPM with 2-4 fingers lol

I program with the German layout, found it pretty okay.

I found it ok until I switched to UK layout. It takes a while to get used to the layout for programming but it does speed up programming a lot for languages that use brackets excessively. Others (e.g. Python) don't really make a difference.

Was not the fastest typer in the world a Dvorak user? http://rcranger.mysite.syr.edu/dvorak/ Not exactly a study but it can't be ignored in my opinion. Disclosure I am a Dvorak user so I may be biased.

I get what you're saying, but max(f(x)) isn't a very good way to characterize a sampling distribution. It could be that for very few people, Dvorak is blazingly fast while for the vast majority of experienced typists there is zero (or negative) improvement.

I have no idea if Dvorak is better or not, but this data point isn't convincing.

> but max(f(x)) isn't a very good way to characterize a sampling distribution

That depends greatly on what your goal is. I would appreciate it if max (or something close to it) was the most common for comparing frame rendering latency (inverse of frame rate) measurements. The large number of fast QWERTY typists could be due to the greater number of overall QWERTY typists, who knows.

This is not relevant. Frame rendering latency matters because as an end user this effects performance in an obvious and measurable way. The thing you care about depends on the worst case, so it makes sense it that case. I don't type faster because a competitive typist hit a certain wpm.

Was not the fastest typer in the world a Dvorak user?

150 wpm for 50 min (37,500 key strokes) and attained a speed of 170 wpm using the Dvorak Simplified Keyboard (DSK) system. Her top speed was recorded at 212 wpm.

No. http://seanwrona.com/typing.php

50 minute typing in English - 174 wpm (870 cpm) - 1st all time (this unofficially breaks Barbara Blackburn's Guinness world record of 150 wpm (750 cpm) over 50 minutes, but is perhaps less impressive because she used a typewriter while I used a computer keyboard).

Caveat aside, his peak speed is over 250WPM.

If you look at typeracer.com's scores, there are plenty of typists in the 150+ range. I'd also consider myself a fast typist, averaging 150-160 for general English text and have peaked at ~240wpm in short bursts. I use QWERTY exclusively, but on an extremely low-force keyboard.

Reply on news.ycombinator.com