Dvorak keyboard layout, anyone used it?

ascii · Post by **ascii** » Fri Sep 24, 2004 6:52 am

wim: If you read throgh my comments I have never claimed that the Qwerty keboard is better than the Dvorzak layout. I've repeatedly said that studies show very little difference in typingspeed between the two. And if you read throgh the articles I've linked, you will find tests showing this deifference to be between 2% and 6%, both in favour of the Dvorzak keyboard.

In your post you go to great lengths to prove that the Qwerty design isn't better at alternating hands than Dvorzak, which is something I have never claimed. You asked me for words that are hard to type in Dvorzak since you said "i challenge you to find an equally awkward word on the dvorak layout". If puppylike or seppuku is harder to type on Dvorzak than lollipop is on Qwerty is subjective, but according to the speed guidlenes about avoing using only one or two fingers when typing, I think we can agree that I provided you with a rather lengthy list of hard to type words. Reading though your test results, I find that the Qwerty keyboard has an alternation frequency of about 25 to 30%. If no care had been taken to maximize alternation, this frequency should be around 50%. Clearly you have just proven that Qwerty was designed using principles wich lead to a higher than average hand alternation. Which is exactly what I originally said.

Yes, Dvorzak is a better designed keyboard layout. I do not claim the opposite. I claim:
1). That Qwerty's keyboard layout isn't so bad as people make it out to be, which you yourself helped me prove.
2). The difference in typing speed between Qwerty and Dvorzak is small.

As to your question of "how can you claim this as a myth without having even bothered to try and disprove it to yourself? you could never be a real scientist.", I suggest you start reding things more carfully and not jumping to conclusions. That leads to bad science.

wim · Post by **wim** » Fri Sep 24, 2004 7:41 pm

ascii wrote:wim: If you read throgh my comments I have never claimed that the Qwerty keboard is better than the Dvorzak layout.

you've claimed that qwerty is not significantly worse, which i'd like to convince you it is. as others have pointed out to you, the comfort and ease of typing is the most noticeable benefit of using the layout, and you have completely ignored this factor in your measure of which keyboard is 'better'.
the only factor you seem to have taken into account is speed, but by that argument we should all be using keyboards shaped like the typewriters they use in courts.
(let me tell you one thing for sure, if the dvorak layout wasn't so much better, i wouldn't still be using it!)

you've also made spurious claims such as

The Dvorak keyboard lets the fingers stay at the home row a lot, which is good for speed, but has a slightly worse spread between fingers, meaning that you wont be able to use all the fingers of your hands as effictively while typing as a Qwerty keyboard.

which you leave unjustified and ungrounded.

In your post you go to great lengths to prove that the Qwerty design isn't better at alternating hands than Dvorzak, which is something I have never claimed.

you claimed qwerty favours Conditiion C, "The frequency of alternating hand sequences is maximized and the frequency of same-finger typing is minimized." and you also say "the Qwerty layout is excelent at minimizing both same-hand and same-finger typing"

you're clearly contradicting yourself here. do you know what "maximized" means?
i could write a loop to disprove the claim about same-finger typing aswell if you'd like, because i don't believe that either.

You asked me for words that are hard to type in Dvorzak since you said "i challenge you to find an equally awkward word on the dvorak layout". If puppylike or seppuku is harder to type on Dvorzak than lollipop is on Qwerty is subjective, but according to the speed guidlenes about avoing using only one or two fingers when typing, I think we can agree that I provided you with a rather lengthy list of hard to type words.

i don't think we can agree. seppuku is a japanese word, the dvorak layout was optimised for english. quippu is not even a word according to dictionary.com. puppylike, okay, but it's just as bad on qwerty. as for these ones:

okupukupu, kukupa, Kippy, Bifid, biddy-biddy

well the only reason i didn't reply was because i thought you were joking. i have never heard of ANY of those words. fwiw, the last two are actually particularly easy ones on the dvorak layout.

if you think you can keep it up, i'll challenge you again to try and match this list:
aftereffects, reverberated, greater, opinion, regard, desegregated, monopoly, devastated, exacerbate, exaggerated, stargazers, streetcars, sweetbread, aftertaste, uphill, killjoy, million, minimum, pumpkin...

Reading though your test results, I find that the Qwerty keyboard has an alternation frequency of about 25 to 30%. If no care had been taken to maximize alternation, this frequency should be around 50%.

this is incorrect.
the presence of spaces in normal typing (about 30% of adjacent pairs of characters in normal text contain a space) means that expected alternation frequency should not be up around 50% but actually a lot lower.

i ran the test again, this time instead of counting total characters i counted pairs of characters which could be an alternation (i.e. count only pairs of letters, not spaces). this resulted in 6015 out of 19101 same-hand presses for dvorak (31%) and 9254 out of 19101 same-hand presses for qwerty (48%).

Clearly you have just proven that Qwerty was designed using principles wich lead to a higher than average hand alternation. Which is exactly what I originally said.

you said it was maximised, not that it was "higher than average"

in any case, the qwerty layout only gets hand alternations between letters about 52% of the time! in a random layout, you'd expect 50%. while 52% is higher than average, you could hardly say that was purposely caused by design principles. that 2% is well within a margin of random 'luck' for the qwerty layout.

As to your question of "how can you claim this as a myth without having even bothered to try and disprove it to yourself? you could never be a real scientist.", I suggest you start reding things more carfully and not jumping to conclusions. That leads to bad science.

speak for yourself. what you're suggesting i do is exactly how you failed when you claim 'an alternation frequency of about 25 to 30%' - erroneously - on which you base claim 1). and you dare say i helped you 'prove' this!?

i didn't jump to conclusions after reading some one-sided articles, i got in there and tested it myself.

finally, why are you calling him "Dvorzak" now?

ascii · Post by **ascii** » Sat Sep 25, 2004 3:11 am

Wim: Yes, I've claimed that Qwery is not significanlty worse, speedwise. I still claim that Qwerty is not significantly worse. You keep bringing up all the theoretical advantages of Dvorak as proof of it's superiority, when there are numerous tests (not just the Strong tests, read my previous links) showing the difference to be minimal. You say you do not belive in studies, but you cling to your theoretical ideas on why Dvorak ought to be better. Newsflash: If reality does not conform with theory, reality is usually right.

And yes, the only factor I've taken into account in my disscussion with you is speed. I have repeatedly statetd that I don't know if the Dvorak keyboard is more or less comfortable, causes more or less strain on your wrists etc., etc.. I am only trying to debunk the myth that switching to Dvorak makes you a much faster typist. If you want to be a faster typist and have already learned the Qwerty keyboard, your time is better spent practicing Qwerty than switching to a new (and only slightly better) layout. That is the essence of what I am trying to say. As to those typewriters used in courtrooms, I'm not sure what type you are talking about. There are special high speed typwriters which instead of working with letters mostly have wordparts on every key, and an insane number of keys. They would be useless for things like writing code (with lots of mathematical operators and such), and if I'm not mistaken they have diffrerent layouts for different languages, depending on what combination of letters are the most common. I don't think that they would be worth the effort, just as I don't think Dvorak is worth the effort.

Reading throgh my comments I can see that the quote you provide about worse spread is actually ungrounded and probably wrong. I was jumping to conclusions. I do not think this invalidates what I have otherwise said.

And I do not understand how I am contradicting myself. Minimizing same-hand typing means exactly the same thing as maximizing hand alternating frequency. Same goes for fingers. Or did you mean something else?

All the words I posted are part of 'web2', which is the most common (AFAIK) wordlist used in Unix. It is based on Websters dictionary. I admit I hadn't heard of some of the words in my second post either, but if you would have taken the time to google them you would have found that they are real. Okupukupu is a kind of plant, kukupa is a bird and so on. I didn't include them in my first post because they where so strange. Seppuku is a word borrowed from japanese. Just as is you read a dictionary, most words in the english language are borrowed from one language or another. Only difference is that seppuku is a rather new word. Makes it no less of a valid word, IMO. And puppylike uses a lthe right hand much more than the left on Qwerty, but it uses all four fingers of the right hand. This is all subjective and irrelevant, but IMO it's much easier to type on Qwerty than on Dvorak.

And finding a few strange words that are hard to type in either layout doesn't prove anytthing. Real world typing speed is what matters. You originally DARED me to find any words that could match lollipop as a hard word to write, which is the only reason I tried to find them. You admit that puppylike is such a word. But instead of confessing that you where wrong in thinking no such words existed for Dvorak, you immediatly challenge me to find more words. I could probably find a few more hard to write words, but that sounds rather boring, I must confess. I admit that you would probably win, because Qwerty is a worse layout than Dvorak. The point which I was trying to get through to you is that Dvorak isn't as amazing as it's been made out to be. It has words that are hard to type, just like Qwerty.

Since you didn't provide code for your altered calculations I will need to write my own code to check your claims. I'll see if I am stubborn enough to do this, but I can't comment on those claims until I have done so.

About the difference between maximizing and 'higher than average', neither Qwerty or Dvorak can maximize any such meassure in the strictest sense, because language changes, different people use different language, etc. I should probably not have used the term maximize because it can imply reaching a theoretical maximum. What I meant with maximize was simply try 'to reach as high as possible'.

Oh, I misspelled Dvorak. Why did you feel the need to point that out? I belive this is rather bad netiquette.

sthayashi · Post by **sthayashi** » Sun Sep 26, 2004 4:44 am

ascii wrote:Oh, I misspelled Dvorak. Why did you feel the need to point that out? I belive this is rather bad netiquette.

Actually, I was curious about this because you were consistent on this mistake in your last post (a single instance can be an error of the fingers, multiple instances of the same error suggest that the error was in the mind

).

Wim, since you seems to be in the habit of writing a program to run calculations for you, perhaps it would be worthwhile to calculate the number of words that deviate from the home row, with emphasis on letters and characters that REALLY deviate like T,Y,B,N on Qwerty and Y, F, X, B on Dvorak (like 1 point for non-home row keys, 2 points the characters I just listed). One could call this informal test the "Finger stress test," with the idea that the more you move your finger, the more it's stressed.

I remember when learning the Dvorak system was that it was a lot easier to become more useful after only learning the home row set. The only downside was remembering UNIX commands in Dvorak.

alleycat · Post by **alleycat** » Sun Sep 26, 2004 5:50 am

Isn't "Dvorak" correctly pronounced as "Dvorzak"?

Spod · Post by **Spod** » Tue Sep 28, 2004 7:15 am

Yes, but Dvorak is the correct approximation of his name when restrictricted to the English character set - the "z" sound comes from some sort of accent over the r or the a.

As for the recent debate, it seems that the main reason to choose the dvorak layout is comfort of typing, with speed coming with practise.
If someone finds dvorak more comfortable, who cares what the statistics and minutiae of the design are? If it proves uncomfortable, just stop using it.

pangit · Post by **pangit** » Tue Sep 28, 2004 6:16 pm

alleycat wrote:Isn't "Dvorak" correctly pronounced as "Dvorzak"?

Yes if you're referring to the composer, no if you're referring to the keyboard layout.

wim · Post by **wim** » Sat Oct 02, 2004 3:32 am

Spod wrote:Yes, but Dvorak is the correct approximation of his name when restrictricted to the English character set - the "z" sound comes from some sort of accent over the r or the a.

err, no, it is pronounced how it looks - duh-VOR-ack - not like the Czech composer

alleycat · Post by **alleycat** » Sat Oct 02, 2004 4:26 am

Looks like another one of those 'manglicized' words.

wim · Post by **wim** » Sat Oct 02, 2004 4:40 am

finally found the time to compose my proper reply in this puppylike thread

ascii wrote:And I do not understand how I am contradicting myself. Minimizing same-hand typing means exactly the same thing as maximizing hand alternating frequency. Same goes for fingers. Or did you mean something else?

yeah i did mean something else (my bad, i read you wrong!

)
you did not contradict yourself - but you're still contradicting me which i don't like.

if you'll at least agree/accept that the loop shows hand alternation in qwerty is significantly worse than other layouts can make possible, then you'll be contradicting yourself (so you'd better find an error in my code..

)

Dvorak isn't as amazing as it's been made out to be. It has words that are hard to type, just like Qwerty.

because of the sheer number of words that exist in english, way WAY more than most other languages, i'd wager it would be literally impossible to arrange the keys in ANY way so as to avoid existence at least a couple of long one-hand words (and i'm amazed that dvorak does as well as it does here - yep it is amazing, i think

). ok - so the sensible aim is to try harder to avoid those which are typed more often. i don't describe things as 'puppylike' all that often, and okupukupukupukku kukupapakupablablabla i'd never typed before in my life.

Since you didn't provide code for your altered calculations I will need to write my own code to check your claims. I'll see if I am stubborn enough to do this, but I can't comment on those claims until I have done so.

it's good you're skeptical enough to verify my claims, if only you would do the same for the references you use earlier. as for the code addition - ffs it's so trivial i didn't think i'd have to bother to mention the code. just add a counter (count = count+1) inside the first 'if' statement, and use this value [count] for your ratios rather than the number of characters [length(str)]. and, by the way, only you made that mistake in the first place - i did not 'alter' my calculations, i was correcting your wrong calculations.

About the difference between maximizing and 'higher than average', neither Qwerty or Dvorak can maximize any such meassure in the strictest sense, because language changes, different people use different language, etc.

the issue here is that how 'maximised' it is depends on the sample text, so this measure is poorly defined - you would have to test it against all text ever written. but it's much easier to compare two layouts against eachother and say this one does better than that one, with arbitrary certainty. and you'd probably have to create a REALLY contrived sample of text to see qwerty ever beat dvorak in this test.

and it IS possible to reach, say, with the nearest 1% possible for any sufficiently large sample of typical text, which i'd surmise the dvorak layout probably does acheive.

I should probably not have used the term maximize because it can imply reaching a theoretical maximum. What I meant with maximize was simply try 'to reach as high as possible'.

and you think sholes has done this with his typewriter layout..?

Oh, I misspelled Dvorak. Why did you feel the need to point that out? I belive this is rather bad netiquette.

like sthayashi mentioned, i figured since you misspelled his name 7 times, it can't have been a typo and maybe it was some joke i wasn't understanding.
but now that you've clarified that you couldn't even get his name right, it just makes it seem to me that you've barged in hear to spout crap about something you know nothing about. so if you were just looking for an argument, i hope i gave you what you wanted.

cheers!

ps: put my netiquette in yer pipe and smoke it.

alleycat · Post by **alleycat** » Sat Oct 02, 2004 4:52 am

Speaking of bad netiquette... I hear you've done a bit of barging yourself!

wim · Post by **wim** » Fri Nov 05, 2004 5:18 am

here's an interesting read, this guy wrote 1000 lines of C to try and "evolve" the best keyboard layout, using a reward/penalty system (actually..similar to what sthayashi suggests in previous reply) and a few hundred thousand words of input. input including not only just literature but a decade of his emails and C code from his - i presume - job.

the result was a mapping pretty similar to dvorak, and not actually able to significantly beat dvorak by his own measure.

sthayashi · Post by **sthayashi** » Fri Nov 05, 2004 6:01 am

I like it. He even provided a link to generate some statistics between qwerty and dvorak.

fanerman91 · Post by **fanerman91** » Fri Nov 05, 2004 6:44 pm

Here's a reply that's not borderline flaming (just a small update on my Dvorak experience)...

After a little bit of getting used to... I consider myself fairly proficient at both Dvorak and Qwerty. It took awhile to get used to Qwerty again but I think I got it. Switching between the 2 isn't too bad. Though I use Dvorak exclusively at home. I don't really get that option at other places.

sthayashi · Post by **sthayashi** » Sat Nov 06, 2004 11:47 am

fanerman91 wrote:After a little bit of getting used to... I consider myself fairly proficient at both Dvorak and Qwerty. It took awhile to get used to Qwerty again but I think I got it. Switching between the 2 isn't too bad. Though I use Dvorak exclusively at home. I don't really get that option at other places.

This is fairly consistent with my experience. One of the problems you may encounter however, is the feel of the keyboard and position. If you come across a keyboard that feels exactly like what you have at home, and you have in the exact same position, you'll probably find that you'll make a lot more qwerty/dvorak errors.

wim · Post by **wim** » Sat Nov 06, 2004 11:56 pm

fanerman91 wrote:Switching between the 2 isn't too bad.

i hate it. i mean i can still string through a sentence in Q without looking at the keys much, but the hands seem to go flailing about all over the place and makes me feel spastic.

fanerman91 wrote:Though I use Dvorak exclusively at home. I don't really get that option at other places.

have you tried that DVAssist.exe linked earlier? it's quite a treat..

Zyzzyx · Post by **Zyzzyx** » Sun Nov 07, 2004 12:53 am

Since this has come back again, thought I'd toss in an update.

It really doesn't seem like it was way back in January that I changed over. But then it also seems like I've always been typing with this layout. Weird.

Just visited my favorite TypingTest.com the other day, outta the blue. Managed a 54wpm on a three minute test. Looking back to my last post of speed results three months ago in August, I was able to reach 55wpm Dvorak then.

Have I topped out on my Dvorak typing speed? I don't think so. I still seem to have far more corrected errors (well worn backspace key) than I remember having with Qwerty. Topped out? No, probably not. Plateaued? Likely.

mathias · Post by **mathias** » Sun Nov 07, 2004 7:38 am

Guess I'll add my progress report, I've used it for one and a half months, I'm up to 29 WPM, compared to the 35 I used to do on qwerty. The important thing is it feels much more comfortable.

snutten · Post by **snutten** » Mon Nov 08, 2004 8:30 pm

Anybody tried Dvorak for swedish (or other non-english language)?

Straker · Post by **Straker** » Mon Nov 08, 2004 11:27 pm

ouch, 55wpm? 29wpm?

i'd love to try dvorak and have thought about it a few times, but the change would kill me, i already type 120wpm or so using qwerty.

sthayashi · Post by **sthayashi** » Tue Nov 09, 2004 8:25 am

Straker wrote:ouch, 55wpm? 29wpm?

i'd love to try dvorak and have thought about it a few times, but the change would kill me, i already type 120wpm or so using qwerty.

I think it really needs to be brought into comparison with qwerty speed. I know that personally, I still make a lot of mistakes. My backspace key is easily the most used key on the keyboard. And most of it are the same mistakes that I made before (e.g. just now I typed in 'keyborad').

Assuming you can afford the time to retrain, I'd give it a shot. Especially if your wrists have ever hurt from typing too much.

Zyzzyx · Post by **Zyzzyx** » Tue Nov 09, 2004 10:19 pm

Straker wrote:ouch, 55wpm? 29wpm?

i'd love to try dvorak and have thought about it a few times, but the change would kill me, i already type 120wpm or so using qwerty.

55wpm with Dvorak is about right for me. IIRC, I was only hitting upper 60s wpm with Qwerty. And I'm still slowly getting better.

Straker · Post by **Straker** » Wed Nov 10, 2004 4:57 pm

sthayashi wrote:I think it really needs to be brought into comparison with qwerty speed. I know that personally, I still make a lot of mistakes. My backspace key is easily the most used key on the keyboard. And most of it are the same mistakes that I made before (e.g. just now I typed in 'keyborad').

Assuming you can afford the time to retrain, I'd give it a shot. Especially if your wrists have ever hurt from typing too much.

sorry, yeah, i know qwerty's at best not any faster (and i also know that wasn't your point either

), just saying it's a huge sacrifice starting over when i can already type this fast...
then again, not like i need the keys labelled properly to type in qwerty if i need to while learning dvorak... hmm.

Zyzzyx · Post by **Zyzzyx** » Mon May 09, 2005 1:23 pm

Its been awhile, I dunno why this came to mind today, but it did, so I'll toss in an update.

Back in November:

Zyzzyx wrote:Just visited my favorite TypingTest.com the other day, outta the blue. Managed a 54wpm on a three minute test. Looking back to my last post of speed results three months ago in August, I was able to reach 55wpm Dvorak then.

Have I topped out on my Dvorak typing speed? I don't think so. I still seem to have far more corrected errors (well worn backspace key) than I remember having with Qwerty. Topped out? No, probably not. Plateaued? Likely.

Well, looks like that plateau has slowly been raising. Just did another typing speed test and managed a comfortable 70wpm on a three minute test. On the one minute test I could 'sprint' to 78wpm. Again, I'm not sure if I'll get any faster from here, but it could happen.

StarfishChris · Post by **StarfishChris** » Mon May 09, 2005 3:12 pm

I'm thinking of giving this a go after a few months of wondering what the heck it's all about and if it's a good idea to switch. I'm at 88WPM with Qwerty and almost exclusively burst-type - very rarely do I continuously type, and I drop down to about 60 then. The idea of typing one-handed is appealing and would be very useful for the kind of work I do, but would you recommend I learn the left-hand first (if at all) or the standard?

Splinter · Post by **Splinter** » Mon May 09, 2005 4:07 pm

peteamer wrote: There is also at least one other format... but that's too geeky to know about and i'm not going to tell you. Doh!!!.

Well actually I cant recall much about it/them other than they exist.

AZERTY

StarfishChris · Post by **StarfishChris** » Mon May 09, 2005 5:02 pm

AZERTY is a French keyboard (I had the misfortune to try and enter numbers on a laptop version, with no numberpad and not knowing you had to hold Shift for the numbers!)

wim · Post by **wim** » Mon May 09, 2005 5:29 pm

StarfishChris wrote:The idea of typing one-handed is appealing and would be very useful for the kind of work I do, but would you recommend I learn the left-hand first (if at all) or the standard?

i should just point out that one hand typing on the dvorak standard layout is REALLY HARD
(this is of course because it's optimised for alternate hand typing)

StarfishChris · Post by **StarfishChris** » Mon May 09, 2005 6:00 pm

wim wrote:
StarfishChris wrote:The idea of typing one-handed is appealing and would be very useful for the kind of work I do, but would you recommend I learn the left-hand first (if at all) or the standard?
i should just point out that one hand typing on the dvorak standard layout is REALLY HARD
(this is of course because it's optimised for alternate hand typing)

That's why I'm suggesting the use of the left-handed Dvorak keyboard, designed specifically for one-handed typists.

sthayashi · Post by **sthayashi** » Mon May 09, 2005 7:18 pm

StarfishChris wrote:I'm thinking of giving this a go after a few months of wondering what the heck it's all about and if it's a good idea to switch. I'm at 88WPM with Qwerty and almost exclusively burst-type - very rarely do I continuously type, and I drop down to about 60 then. The idea of typing one-handed is appealing and would be very useful for the kind of work I do, but would you recommend I learn the left-hand first (if at all) or the standard?

I'd say the standard, simply because there are more resources, and fewer of the keys have been moved.

For example, in normal Dvorak, the numbers are still the same. In single handed Dvorak, they're moved. Even more so, the style of two-handed typing is still the same (i.e. the concept of a home row and what fingers are meant to press which keys). In single handed Dvorak, it's never been clear to me which keys were meant to be pressed by which fingers.

However, if you like to eat and type at the same time, then maybe a single handed dvorak format may be for you.