geekhack

geekhack Community => Keyboards => Topic started by: thrab on Wed, 12 January 2011, 22:08:51

Title: Workman - Interesting Read
Post by: thrab on Wed, 12 January 2011, 22:08:51: Some good research and keyboard layout. Too bad my muscle memory is so ingrained I will never be able to use anything other than QWERTY :pout:

http://viralintrospection.wordpress.com/2010/09/06/a-different-philosophy-in-designing-keyboard-layouts/ (http://viralintrospection.wordpress.com/2010/09/06/a-different-philosophy-in-designing-keyboard-layouts/)
Title: Workman - Interesting Read
Post by: Viett on Wed, 12 January 2011, 23:03:48: Interesting. It's too bad trying out layouts is not exactly a simple process.

But it's clear that Colemak is by no means perfect. Some of the statistics at Carpalx (http://mkweb.bcgsc.ca/carpalx/) tell the same story.

Colemak truly is just the "popular alternative." At this point it's far better than Dvorak and QWERTY and enough people use it that it can be called a common layout. Trying to overshadow Colemak with a slightly better layout just isn't going to happen. Better to keep the Colemak momentum going in the community than to criticize its shortcomings and start from scratch.
Title: Workman - Interesting Read
Post by: j_r on Wed, 12 January 2011, 23:34:12: Thank you for the post. I really never think about the key layout much since I am relatively quick with the standard layout - and I am relearning how to play the violin and guitar, I do not need another item in my life I have to relearn. If anything I have a problem more with the auxiliary symbols on a keyboard... I was playing around with Windows PowerShell and wanted an easier means to get the "-" and "|" keys. Granted I am still ****ing with a terrible membrane keyboard - but hopefully that will soon come to an end.
Title: Workman - Interesting Read
Post by: Tony on Thu, 13 January 2011, 01:11:58: After 15 years of touch typing in Qwerty (speed 60wpm), I am now learning Colemak layout for 38 days and my speed in Colemak is already 49wpm.

So it is amazing how human brain (and muscle memory) can adapt. To reach the same Qwerty speed again in another layout will takes you about 6 months or 1 year, but you can reach 2/3 of that speed pretty fast.

Colemak (http://Colemak.com), Dvorak, Arensito, Capewell or Workman (http://viralintrospection.wordpress.com/2010/09/06/a-different-philosophy-in-designing-keyboard-layouts/) is quite the same to me. Their difference are too small (1-2%), so you can choose any modern layout you want. It is only a matter of preference.

All of them are 30% superior to Qwerty according to key travel. With Qwerty, my fingers are flying constantly in the top row, but with Colemak, my fingers are resting in the home row most of the time.

My wrists are free of pain and I can type longer.

Edit: Some layout pictures for anyone who wants to know

(http://img705.imageshack.us/img705/1112/colemak.png)
Colemak

(http://img266.imageshack.us/img266/8633/dvorak.png)
Dvorak

(http://img831.imageshack.us/img831/1197/capewell.png)
Capewell

(http://img406.imageshack.us/img406/3429/arensito.png)
Arensito
Title: Workman - Interesting Read
Post by: taswyn on Thu, 13 January 2011, 08:12:49: I stopped reading when the bald faced duplicitous statistics misrepresentation began. Here, look:

Code: [Select]
Colemak Distance: 30352m Same hand: 31.17% Same finger: 1.726% Number row: 0.217% Top row: 15.96% Home row: 73.59% Bottom row: 10.22%
Code: [Select]
Workman Distance: 29656m Same hand: 32.99% Same finger: 2.185% Number row: 0.217% Top row: 19.76% Home row: 67.63% Bottom row: 12.37%
Quote
Workman, on average, has a higher SFU than Colemak… at +1%. Some people misunderstand and think that this somehow shows increased effort or discomfort. It doesn’t.

39831 single finger combos versus 50424.

If you really want to exaggerate, you could refer to "key strokes part of a single finger combo" and double this.

Quote
Looking at the first example. Colemak achieves the lowest overall finger-travel distance against QWERTY and Dvorak at 30,352 meters. However, Workman is even lower at 29,656 meters — a difference of 696 meters. It doesn’t sound like much, however if we convert it to centimeters, that’s equal to 69,600 cm. And considering that the distance between keyboard keys is approximately 2 cm, typing on Workman is like typing 34,000 less keystrokes than typing on Colemak. At 40 words per minute, that’s equivalent to approximately 3 hours of work.

or a 2.3% increase from Workman to Colemak (I was generous, since this looks smaller from the other direction).

Is the problem apparent yet? Even if the two deserve a direct and equivalent comparison, it's being done with different metrics and resolutions to completely skew the *perceptive* balance of differences.

And I could use comparisons to Qwerty to both amplify and reduce apparent differences, if *I* wanted to.

I don't use either, so I shouldn't even really care, and maybe it's just an innocuous case of someone getting excited about their work with no malicious intent to mislead or anything, it just is a huge pet peeve of mine when people don't make an effort to keep their statistical comparisons honest by sticking to consistent comparisons.

There may be some good things about Workman, even better than Colemak in certain regards, but things like this really turn me off to much more exploration. As a persuasive paper, it's simply too disingenuously biased for me to expect a sincere comparison of faults and strengths.

For instance, a quick look back over some of the numbers makes me suspect Workman may have a slight RSI advantage, with Colemak probably having a slight speed advantage. But as pointed out, Colemak is already a recognized standard. I doubt Workman can show a significant enough of an improvement to be worth even just that particular difference.

it's all fairly moot to me, since I feel like thumb interface design represents the most significant failure of the modern keyboard. Energy on key layout design seems better focused there than on trying to eek out very mild improvements to something Colemak already essentially solved for current standard keyboards.

On the plus side, this may have convinced me to switch to Colemak from Qwerty =P

... Eventually >.>
Title: Workman - Interesting Read
Post by: Daniel Beaver on Thu, 13 January 2011, 08:53:54: On the subject of Dvorak:
Quote
i went ahead and tried it out and soon enough after doing “ls -latr” on the terminal, i had to shake my head and sadly walk away from it.
It's so true. Why in the hell is the L there?
Title: Workman - Interesting Read
Post by: Viett on Thu, 13 January 2011, 11:04:40: Quote from: taswyn;277828
I stopped reading when the bald faced duplicitous statistics misrepresentation began. Here, look:

...

And I could use comparisons to Qwerty to both amplify and reduce apparent differences, if *I* wanted to.

Funny thing is that I looked at his statistics and concluded that there was little relative difference between the layouts, and almost dismissed it immediately. I didn't read his absurd conclusion -- that that small difference mattered. I guess it's easy to amplify a 2.2% relative difference with a 400,000 word text.
Title: Workman - Interesting Read
Post by: Tony on Thu, 13 January 2011, 20:29:41: For a keyboard layout to stay alive, efficiency is a small factor.

Numbers can only impress technicians and mathematicians. Familiarity may impress other people who do not want to invest much time leaning new layout. Or they don't type much to justify the change.

Other factors like community, public press, and good feedback from users should be taken into account.

We still remember Dvorak is a much more efficient layout than Qwerty, yet Qwerty control the market until now. To move from Qwerty to Dvorak is too radical, they change 33 keys.

(http://img842.imageshack.us/img842/9264/colemak2.jpg)

Colemak (http://Colemak.com) change only 17 keys from Qwerty, and it is the reason I decide to switch, after 15 years touch typing in Qwerty.
Title: Workman - Interesting Read
Post by: msiegel on Thu, 13 January 2011, 21:22:21: Quote from: Tony;278261
Colemak change only 17 keys from Qwerty

better "backward compatibility" *and* excellent efficiency is a very good combination :)
Title: Workman - Interesting Read
Post by: Tony on Thu, 13 January 2011, 23:11:19: Quote from: msiegel;278289
better "backward compatibility" *and* excellent efficiency is a very good combination :)

Exactly.

And Colemak (http://Colemak.com) is not a fashion, you need to use the keyboard more than 2 hours a day to justify the layout change, due to the learning curve.

From Qwerty to Colemak, there's 30% improvement, so it is worth it if you type a lot. For game, I still use Qwerty layout. Just press a hotkey that can be set up in the Control Panel.

From Dvorak, Arensito, or any modern layout, there's only 1-2% improvement, so stick with your current layout and don't convert.
Title: Workman - Interesting Read
Post by: Lanx on Sat, 15 January 2011, 22:38:13: Quote from: Tony;278261

Show Image
(http://img842.imageshack.us/img842/9264/colemak2.jpg)

Colemak (http://Colemak.com) change only 17 keys from Qwerty, and it is the reason I decide to switch, after 15 years touch typing in Qwerty.

oh this pic is cool and makes me want to try colemak (tho i'm colorblind so the non blue colors all look the same) and colemak keeps the x,c,v as the same which i think is essential to normal usage.
Title: Workman - Interesting Read
Post by: hoggy on Tue, 18 January 2011, 16:16:39: I thought the recommendation of a typematrix was a nice touch...
Title: Workman - Interesting Read
Post by: Keymonger on Fri, 21 January 2011, 09:18:49: That blog post got me thinking a while ago, and I got around to creating my own layout, using this tool: http://patorjk.com/keyboard-layout-analyzer/

Using it now. So... slow... 10-12 wpm. First impression is amazing. Coming from dvorak, surprised by the improvement. Can't say more though, need more time with it.
Title: Workman - Interesting Read
Post by: eMps on Fri, 21 January 2011, 10:51:56: Quote from: Keymonger;282236
That blog post got me thinking a while ago, and I got around to creating my own layout, using this tool: http://patorjk.com/keyboard-layout-analyzer/

Using it now. So... slow... 10-12 wpm. First impression is amazing. Coming from dvorak, surprised by the improvement. Can't say more though, need more time with it.

Do share the layout with us :)
Title: Workman - Interesting Read
Post by: Keymonger on Fri, 21 January 2011, 14:02:29: Keeping it under wraps for now, for simple reason, I designed it with my own fingers in mind and may not work well for others... will have more to say when I've hit 50 wpm. Did test just now... 9 wpm :ballchain:
Title: Workman - Interesting Read
Post by: Tony on Sat, 22 January 2011, 03:53:05: Each keyboard layout would requires at least 2 months of practice to get truly comfortable using it.

The switch experience is quite hellish, difficult and exasperated to anyone, so I suggest you should not do that often. One or two for a lifetime would be enough.

Of course, if you have plenty of free time and need to do something that keep you occupied, do your thing.
Title: Workman - Interesting Read
Post by: JBert on Mon, 31 January 2011, 16:34:09: It just optimizes away some of the requirements Shai Coleman set out when he started on the Colemak layout for some other subjective requirements. Here's what Coleman says about the Workman layout (http://forum.colemak.com/viewtopic.php?pid=6842#p6842).

Everyone is allowed his opinion and by extension his keyboard layout requirements, yet I still think the Colemak makes a whole lot of sense.
Title: Workman - Interesting Read
Post by: Keymonger on Tue, 01 February 2011, 10:05:51: Quote from: JBert;287926
It just optimizes away some of the requirements Shai Coleman set out when he started on the Colemak layout for some other subjective requirements. Here's what Coleman says about the Workman layout (http://forum.colemak.com/viewtopic.php?pid=6842#p6842).

Everyone is allowed his opinion and by extension his keyboard layout requirements, yet I still think the Colemak makes a whole lot of sense.

I am utterly unimpressed by those criticisms. I think it's bad to assume everyone types the same way. For example, "6. It has more row jumping than Colemak." Big whoop. I feel that row jumping is irrelevant.

How 'effort' is measured on some layout analysis websites is also a problem I think. Distance is a good measure; but as I've found out, all fingers are not equal, and distance traveled by the pinky is not the same as distance traveled by the index finger. Even between the index and middle finger, the difference is noticeable.

Shai's final critique is just plain silly:
Quote
11. A project that only exists as a blog post, and doesn't even have it's own webpage doesn't inspire too much respect, or demonstrates any investment from the author.

Apparently doing a keyboard layout is so demanding that you need a lot of researching to find something good. That's not true, and I'll have more to say about that later. And there's certainly nothing wrong with promoting a layout using a blog post.

I do think Colemak is the best alternative to Qwerty though.
Title: Workman - Interesting Read
Post by: Tony on Wed, 02 February 2011, 01:10:28: Ideally we would have one keyboard layout for each individual, after calculating each finger's length and muscle strength.

Till then, Workman, Colemak, Dvorak, Arensito, Capewell are all suitable candidates to replace Qwerty.
Title: Workman - Interesting Read
Post by: notlofty on Wed, 16 February 2011, 00:38:58: A few years ago when I first started college and was confronted with the fact that I'd have to do loads of typing I looked into different keyboards and different keyboard layouts (this is when I first found geekhack) so as not to overly strain my already strained hands (former North American Record for solving the Rubiks cub with one hand). Mechanical keyboards cost money so I still don't have one but I switched to Colemak as its free.
I type at around 70-80wpm on it and really like it a lot. When typing the fingers really do stay on the homerow the entire time. But after reading this article about Workman I'm almost tempted to switch, and reading from Keymonger I'm almost tempted to make my own layout. Due to my one handed rubik's cubing and being left handed my left hand is much more dexterous than my right hand is. And while I do agree that moving the fingers up is easier than moving left and right, its really not that big of a deal as he makes it out to be.
I'm very happy with Colemak now but I would be open to switching again if one layout was really shown to be better. And since I never learned to type on Qwerty learning curve from that is irrelevant to me.
Edit: On second thought. I'll probably just stick to Colemak lol. It's done me no wrong thus far.
Title: Workman - Interesting Read
Post by: Tony on Wed, 16 February 2011, 06:13:07: For any layout there will be advantages and disadvantages (pros and cons).

You cannot optimize all options: hand alternative, keyboard rolls, pinky usage, finger travel, home row usage, qwerty-compatibility. If a layout claims to be better in one aspect, it must be worse in another.

Therefore, better stick with the modern layout you already choose (Dvorak, Colemak, Workman, Arensito) since the learning cost is quite high in term of time and effort. You don't have to relearn full keyboard keys for 1.7% improvement.
Title: Workman - Interesting Read
Post by: Keymonger on Wed, 16 February 2011, 11:22:42: Quote from: notlofty;295902
A few years ago when I first started college and was confronted with the fact that I'd have to do loads of typing I looked into different keyboards and different keyboard layouts (this is when I first found geekhack) so as not to overly strain my already strained hands (former North American Record for solving the Rubiks cub with one hand). Mechanical keyboards cost money so I still don't have one but I switched to Colemak as its free.
I type at around 70-80wpm on it and really like it a lot. When typing the fingers really do stay on the homerow the entire time. But after reading this article about Workman I'm almost tempted to switch, and reading from Keymonger I'm almost tempted to make my own layout. Due to my one handed rubik's cubing and being left handed my left hand is much more dexterous than my right hand is. And while I do agree that moving the fingers up is easier than moving left and right, its really not that big of a deal as he makes it out to be.
I'm very happy with Colemak now but I would be open to switching again if one layout was really shown to be better. And since I never learned to type on Qwerty learning curve from that is irrelevant to me.
Edit: On second thought. I'll probably just stick to Colemak lol. It's done me no wrong thus far.

If you like sticking with the home row, Colemak is excellent. That's what it's designed for. But I think keyboard layouts should be designed for what's most comfortable, and I don't quite agree with Colemak's focus on the home row. Specifically, it really seems to avoid upper row movements for middle fingers, which happens to be my favorite. The layout I made is statistically, markedly different compared to other layouts. I have made it so that it completely overloads the middle fingers; they take around 40% of the share when typing in English. Dvorak and Colemak are around 26-29%. The difference is very noticable, and I'm happy to learn that burdening the middle fingers with such a load allows for some very nice, swift movements in English. (Haven't tried typing in Dutch yet.) It seems that all other layouts (that I am aware of), Colemak, Dvorak, Capewell, Workman, try to divide the load among fingers somewhat evenly, which I think is nonsense.

I've been thinking a lot about keyboard designs and layouts for the past few months, and I'm slowly getting to the point where I can be satisfied. But first, I'll need a board similar to the TrulyErgonomic (what a name). I have some very exciting ideas I can't wait to implement.

Using my layout for over three weeks now, I'm at around 40-50 wpm. Using Dvorak, I thought I was at around 70-80 wpm, but looking at some typing test results I did a few months back it seems I was actually around 55-70 wpm. And that was after having learned Dvorak for 6 months, so I wonder how fast I will be. I don't have the exact stats at hand, but my layout's pinky usage is also very low, something like 6% or 8% for both fingers, which I think will help in speed and comfort.

(thread hijack imminent)

Of course I needed a name, and I've decided to call it the Medius layout, because it's focus is on the middle fingers (digitus medius), and that's what sets it apart:

(http://i53.tinypic.com/2whhxnc.png)

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

0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,42,16,23,24,15,43,48,47,20,39,25,26,27,28,31,29,19,18,32,49,30,17,22,21,38,40,41,51,45,44,34,46,33,37,50,35,36,52,53,54,55,56,57,58,59,60

The point in favor of a custom layout is easily made, and the case for moving away from Qwerty is easier still.

If you're using the Qwerty layout, the home row buttons on the left hand are ASDF. First, try typing A, S rapidly. So you'd get asasasasasasasas, etc. Then type D, F repeatedly in succession: dfdfdfdfdfdfdfdfdfdfdf. If you're like me, typing df a lot with your middle and index finger is much easier and faster than typing asasas a lot. Basically, there are movements like that, that your hands are good at, all over the keyboard when you're typing. I'm no expert on this, but this seems to me rather simple.

Taking Qwerty again, doing this:

s d f s d f s d f s d f s d f

is much easier than doing it the other way around:

f d s f d s f d s f d s f d s

Using my own layout, I have noticed many such quick movements that go from ring finger, to middle finger, to index finger, which made it very fun to type when I was learning this layout. (Still, it was a painful experience like Tony says.) I'm a little unsure how I managed to pull this off, but perhaps it is because I overloaded the middle fingers that it allows so many movements of that kind. Or maybe it's just dumb luck. And by the way, even though the middle fingers by far do most of the work, it doesn't feel that way at all. It feels quite balanced. Another interesting point is that L, S are both keys for the pinky on Dvorak. On Medius, they're on the index finger.

I didn't design Medius for anyone else in mind. I saw that Colemak tries to differ from Qwerty as little as possible to allow easy learning, but I didn't like that. I don't want to be burdened by the dead past, so I didn't care about difference to Qwerty at all. There are also some minor considerations. For example, the Unix movement keys HJKL aren't in such a bad position as on Colemak. JK are next to each other and that makes things a bit easier for the Vi users among us (including CLI apps that use Vi-like keys).

I can imagine a piece of software that tells you to type stuff on a keyboard, non-words like as as as and df df df, see how well you do, and suggest a layout based on the collected data sample. The question is, is it worth the trouble. I'm not quite sure. Medius was designed for myself, and I didn't set out to create a layout for the English speaking world. But there might be people out there who would appreciate some extra efficiency. I would be one of those people, but I already took the plunge. It's also possible that you would have a group of layouts. Say, a layout like Colemak, a layout like Medius that focuses on the middle finger, and so on. So, you wouldn't need a totally custom layout, but instead an already existing layout would be suggested to you. And if you happen to be unique, or if there's something wrong with any of your fingers, you could opt for a custom layout.

Just the layout isn't enough. At least, not for me. Standard keyboards are total, asymmetrical mess. So, I have ideas that go beyond mere re-arranging of letters.
I don't think anyone enjoys using shift keys for sustained capital letters, which is why there's a caps lock. But this seems so poorly thought out. You have a modifier, called Shift, that gives you the second layer of keys, that makes letters capital letters. There's a problem, though, because the Shift key is accessed using the pinky. The pinky already covers several keys on the left and right. Which means, if you need a capital A, and the Shift key were on the left, you wouldn't be able to hit them both. So, you have two Shift keys; if you need a capital A, you use the Shift key on the right. But sometimes (or rarely for some people), you need ALL LETTERS to be capital letters. The hands alternate pressing down the Shift keys, and doing that constantly is not pleasant, so there's a third key: caps lock. To accomplish one thing, three keys are necessary. That seems awfully dumb to me. Even worse, the position of caps lock is closer and easier to reach than the far more often reached left Shift key. Just, wow.
I think you can easily reduce those three keys to just one key by having the Shift key accessible by either thumbs. If the space bar was the shift key, it would handle all those things; two Shift keys and caps lock. If you hold down one of your thumbs on the edge if your keyboard's case and pretend for a moment you're pressing a Shift key there, it still allows for great movement by all eight fingers. Using your other thumb, you'd continue hitting the space bar like you normally would. Try typing like that, and it may take some getting used to (my right thumb usually doesn't do anything at all), but I would guess that you'd get used to it pretty quickly. And because it's so easy to hold down and do much typing, it can easily double as a caps lock.
One reason why I think thumbs are underused is that the hands are too close together on a regular board. The thumbs can bump into each other. Separating the hands is also happens to be more comfortable. I think the case for hands needing to be more spaced apart is thus easily made.

There's some more ideas I have, but I've rambled on long enough already. First, I'll need to put some of these ideas into practice, which will hopefully be soon.
Title: Workman - Interesting Read
Post by: Tony on Thu, 17 February 2011, 04:30:16: It's great to see your own keyboard layout. I see that you are happy with it, and that's all that matters.

One question: Why you put U in the home row, which is less frequent than O?
Title: Workman - Interesting Read
Post by: Keymonger on Thu, 17 February 2011, 06:33:31: Getting the middle fingers to do a lot of work was tricky, so I put O there and U on the home row. Result: more middle finger work, less work for the pinky.
Title: Workman - Interesting Read
Post by: Keymonger on Sat, 12 March 2011, 06:42:34: Oh boy. I woke up this morning having an idea without even thinking about. I may have to design a new, more radical layout. At first I wanted to master Medius first but now it seems this idea is so good I need to work on it ASAP. It'll probably use Medius as a base.
Title: Workman - Interesting Read
Post by: hoggy on Sat, 12 March 2011, 11:04:20: I tried the always shifted number row idea out in one of layouts a while back. I type more of the punctuation characters than the numbers themselves so it looked like a good idea.

Couldn't get used to it - I only had it on one keyboard and I swap at least once a day. I think if got used to it I wouldn't go back. It's just too long a process to get used to it.
Title: Workman - Interesting Read
Post by: dsjbirch on Sat, 02 July 2011, 14:58:08: What was the brainwave, Keymonger?
Title: Workman - Interesting Read
Post by: sordna on Sat, 02 July 2011, 19:35:33: Question to colemak users: how easy is it to go back and forth between QWERTY and colemak? I go back and forth between Dvorak and QWERTY, and I have a feeling the fact they are so dissimilar (only 2 keys, A and M are in the same location) makes it easy to distinguish them without confusion. I'm wondering if colemak users can switch back and forth with QWERTY easily.
Title: Workman - Interesting Read
Post by: Tony on Sat, 02 July 2011, 22:36:37: Colemak shares QWAHZXCVBM keys with Qwerty, so switching back and forth is no problem. You only have to differentiate the rest.
Title: Workman - Interesting Read
Post by: Input Nirvana on Sun, 03 July 2011, 01:47:55: I have been using QWERTY on my laptop and Colmak on my Kinesis...seems ok, but I so much prefer Colemak + Kinesis hands down. When I type QWERTY on the laptop, I don't touch type properly.
Title: Workman - Interesting Read
Post by: Playtrumpet on Mon, 15 August 2011, 22:05:34: I think from an objective point of view, it's pretty obvious that there is no outright winner for best keyboard layout, and there may never be. Some people like different sets of movements for their fingers and well researched layouts come into competition and debate that end in taste preference.

For example, some people hate using the bottom row (Dvorak wins there), yet some people don't like using their pinkies so much (Dvorak loses there). For a layout to be in competition, it has to make use of 2 letter pairings (bi/digrams), use comfortable fingers for the most common letters AND digrams (maybe even take into account trigrams if they come up often enough to be relevant), and there shouldn't be many same finger strokes (Colemak does this well), or same hand strokes (Dvorak does this well) (Though, maybe some people don't mind these things or adapt to them OR outright prefer these things, so they may not always be benefits). Honestly, I don't think it's possible to create a layout that can speak best to ALL of these factors (and others not mentioned) and not come short anywhere. THAT is why it really does come down to preference - which factors are most important to YOU.

For me, I really never liked using the bottom row a lot, and I'm fine with using my pinkies. Since Dvorak fit more of my preferences, I decided to try it out rather than Colemak or the other lesser known layouts. I wouldn't turn back at all because I am more comfortable with this layout than I ever thought I would be. I found my layout, but my layout has flaws that some take harder than others, and benefits some people don't consider as important when they type, so it's not for everyone. I encourage people to research a LOT before learning a layout because when it comes to the best layouts in existence, it's like picking your favorite flavor of ice cream.. Even if it's weird, you can't deny that it's best for you if and when you find it.
Title: Workman - Interesting Read
Post by: AUAnonymous on Tue, 16 August 2011, 02:23:49: Not to be that guy too much, but according to Carpalx (http://mkweb.bcgsc.ca/carpalx/?colemak), both Dvorak and Colemak have a 9% bottom row usage rate, although Dvorak does still win when it comes to hand alteration.

I do agree, however, that it's not possible to find the perfect layout and that compromises have to be made regardless of layout.
Title: Workman - Interesting Read
Post by: Proword on Tue, 16 August 2011, 03:08:33: http://proword-keyboardlayoutefficiency.blogspot.com/

Joe
Title: Workman - Interesting Read
Post by: Playtrumpet on Tue, 16 August 2011, 09:31:21: Quote from: AUAnonymous;399737
Not to be that guy too much, but according to Carpalx (http://mkweb.bcgsc.ca/carpalx/?colemak), both Dvorak and Colemak have a 9% bottom row usage rate, although Dvorak does still win when it comes to hand alteration.

I do agree, however, that it's not possible to find the perfect layout and that compromises have to be made regardless of layout.

I noticed that nearly matching 9% bottom row with that list of words, but also noticed that there was almost no punctuation (specifically . , ; and :) used in this list, which (in real writing/typing) makes a much more noticeable percentage difference (I usually see 8-10% for Dvorak, 12-14% for Colemak).

I also wonder what people think about this guy's testing and suggested improved Dvorak called Dvorak (Weak Pinky). http://www.codesharp.co.uk/dvorak/
Title: Workman - Interesting Read
Post by: Proword on Wed, 17 August 2011, 00:49:51: Quote from: Playtrumpet;399835

I also wonder what people think about this guy's testing and suggested improved Dvorak called Dvorak (Weak Pinky). http://www.codesharp.co.uk/dvorak/

Hmm.

I'd probably have to question the methodology. I dropped about 9,000 characters of my current job in there, comparing QWERTY vs Colemak vs Maltron.

In describing Maltron, the test says:

"Maltron layout - this is the layout that comes with the Maltron ergonomic keyboard. It places e on a new key under the left thumb. This essentially creates an extra home row key and results in less effort."

Then in the results it says:

"% Home keys (out of all character keys plus SHIFT and ENTER, but excluding keys operated by thumbs - these are the 8 keys where your fingers normally rest, for example, a s d f j k l ; : on Qwerty)." ??

Saying there are only 8 home keys seems to contradict the above description.

Further down the analysis it says that, other than the space bar, (19.2% of my text), the "e" is the most frequent key in my text at 7.84%, followed by "t" at 7.19%.

One result table says:

_______________Dvorak____Colemak____Maltron
Numbers Row____0.3 %_____1.0 % ______0.3 %
Upper Row______16.3 %____13.0 %____11.7 %
Home Row______53.0 %_____53.8 %____46.1 %
Bottom Row_____11.2 %____13.0 %____13.8 %
Thumbs Row____19.2 %____19.2 %_____28.1 %

So does this in effect mean that the most frequent key in my text is being ignored?

And then at the very bottom of the page:

"Layout Warnings

Warning: The following characters do not exist on layout 'Maltron' and were ignored: E (5 times),"

Joe
Title: Workman - Interesting Read
Post by: sordna on Wed, 17 August 2011, 01:01:53: Well, I do think that calling the E key a home row key is cheating a bit in favor of Maltron. See how much of a skew adding or removing a key makes? The truth is somewhere in between. If I move a couple of letters from the Dvorak layout to my spare thumb keys (like home & page up) on my Kinesis keyboard and call them home row keys (artificially improving the layout stats) I can make my "layout" beat Maltron :-)
Title: Workman - Interesting Read
Post by: Input Nirvana on Wed, 17 August 2011, 01:09:05: Quote from: sordna;400235
Well, I do think that calling the E key a home row key is cheating a bit in favor of Maltron. See how much of a skew adding or removing a key makes? The truth is somewhere in between. If I move a couple of letters from the Dvorak layout to my spare thumb keys (like home & page up) on my Kinesis keyboard and call them home row keys (artificially improving the layout stats) I can make my "layout" beat Maltron :-)

Point well made. I'm in favor of plopping the "E" on the thumb cluster of the Kinesis, I've seen several slightly modified layouts where this has been done, along with a couple modifiers being moved (those damn programmers think they're so special). I would like to see this done to Dvorak and Colemak and see how they compare. I'm not getting too serious about this, but I think it would be fun and interesting.

Quote from: Tony;296000
For any layout there will be advantages and disadvantages (pros and cons).
You cannot optimize all options: hand alternative, keyboard rolls, pinky usage, finger travel, home row usage, qwerty-compatibility. If a layout claims to be better in one aspect, it must be worse in another.
[/FONT][/COLOR][/SIZE]
Title: Workman - Interesting Read
Post by: Proword on Wed, 17 August 2011, 02:59:10: Quote from: sordna;400235
Well, I do think that calling the E key a home row key is cheating a bit in favor of Maltron.

I don't think it's a case of "calling" the E key a "home key". The definition which was in the test said "where your fingers normally rest." It's where the left thumb rests. It just IS a home key. And it's not only in relation to Maltron. As mentioned, there have been other layouts which (could) use the thumbs for something other than space bar.

When I learned to type (QWERTY) back in the mid '60s the thought was (and this still seems to be the case) that the ";" was a home key, yet it doesn't get counted very often in keystroke analysis simply because it's so little used. So there's another skew.

If a keyboard layout can "beat" Maltron by juggling the position of a single key, then that can only be a positive thing which I'd look at.

Joe
Title: Workman - Interesting Read
Post by: Proword on Wed, 17 August 2011, 03:28:46: If you watch this video of me transcribing audio you'll see that my left thumb is pretty well immobile over the "e" key. There's no other place it COULD call home. ;-)

http://www.youtube.com/watch?v=pYJtF1I3PRs

Joe
Title: Workman - Interesting Read
Post by: Input Nirvana on Thu, 18 August 2011, 02:17:15: If nothing else, the Maltron and the Kinesis offer a physical modified keyboard layout that possibly 'creates' an additional home row/key. By being a smarter bear, one can put a high frequency letter in that spot. This will benefit all layouts. I'm thinking that it would most likely improve them all almost proportionately. Maltron wisely chose to create a custom key layout, and to use a letter in the thumb cluster. A Maltron layout can easily be used on a Kinesis, which is a potential cool 'perk'.

One thing I realized about layouts early on. It will almost never matter which one you choose, they are relatively so close to each other once you've made the move away from qwerty, that's where the biggest improvement will be...after that it's just a couple tiny pros/cons that you may personally prefer. And I'm certain there are enough variables that there can't be just one SUPER LAYOUT waiting to be discovered.