"Best" alternative layout

[Sagii]

Hi there, lovely fellas!
I'm terribly sorry if this has been answered countless times before (as it probably has, and if so, feel free to just close this thread), but I'm wondering about DVORAK vs Workman.

So I got all excited when I heard about DVORAK and I started learning it immediatly. However, not long after I saw a comment saying workman is just simply better when thinking about finger movement etc. etc. Essentially my question is this: which one should I learn for just normal typing? I mean, when learning the DVORAK or workman layout you have to put in some considerable time, so I don't want to be "wasted" 'cause there is simply a even better layout.


EDIT: God, I'm an idiot! >_> Literally one thread below mine was an exactly similar one..
Anyways, I might still ask this question: Is it worth learning DVORAK? I'm norwegian, hence I write quite a lot in Norwegian (it is pretty similar to english, taking into considerations which keys are most often used)... I'd say 50/50 norwegian and english. The problem is that as far as I know the only layout available in norwegian is DVORAK. Is the time and effort put into learning DVORAK worth it?

[jakkdl]

As a swede using dvorak, I've decided not to use svorak or svdvorak - and instead remapped alt gr+ueo for åäö. Way better placement of symbols that way.

Regarding dvorak vs workman vs norman vs colemak vs asset vs any other alternative layout, I'd say the differences between them in ergonomics and the like is marginal - but all of them are leagues above qwerty.

Regarding worth, depends on what the goal(s) are, and if can handle the short-term downsides of learning a new layout. Unfortunately qwerty is the standard!

[Vibex]

Currently I use Colemak, but I'm designing my own custom layout. Still needs some fine tuning, but I think it will be quite nice.

[tp4tissue]

If you've already started typing at ~ 40wpm,  then I don't believe going to an "alternative" layout is a good idea..

What you should do, is start basic optimizations of your current layout which you've already partly mastered..

For example.. there's no reason why   colon  should be next to "L "    because that key is super easy to reach and colon is really not important enough to give up on such a valuable position    so for example, you can swap it with the "p "  key..    OR  even the "y " key.

[metalliqaz]

Truly ergonomic?

[Sagii]

Well, I'd say I'm pretty proficient with the QWERTY layout... I do (according to an online test, that is, so don't quote me on that) have a WPM on around 90, so I'm a quick typist. I'm also a touchtypist, if that is of any interest. And the goal with switching to DVORAK would be to get faster, and if it is more ergonomic as well, that's just a plus

[davkol]

Ask Linkbane. ^_^

Or better... don't.

[SonOfSonOfSpock]

As others have said, the difference between the different layouts as far as ergonomics is marginal. Colemak was made so you could switch from qwerty rather easily. I think a lot of it is personal preference.

[jacobolus]

And the goal with switching to DVORAK would be to get faster, and if it is more ergonomic as well, that's just a plus

Switching to Dvorak layout is unlikely to make you super dramatically faster than your current speed, unless you put incredible amounts of practice in, in which case you might get up to 20-30% faster. Indeed, for the first several months, you’ll be slower (initially a lot slower).

The main point is to make typing more comfortable and put less stress on your fingers. With practice, an alternative layout should also slightly improve your efficiency though, and maybe reduce your error rate.

To the same ends, there are several other improvements you can make, if you’re willing to go down the rabbit hole. If you adopt a new physical keyboard layout which is more comfortable for your hands (or even one made from scratch to your personal preferences), and learn to program your keyboard firmware to reorganize the key layout / other functionality, you can end up with something exactly how you want it.

My own recommendations are: load up on thumb keys, make sure the physical layout is easy to type on with your wrists in as neutral as possible a position, and ditch any keys which aren’t completely trivial to reach. Then only after you have a physical layout you like, figure out what kind of character layout to use. (But this is definitely not the easy path.)

Alternately, wait until these guys start taking pre-orders: http://blog.keyboard.io

[Tony]

Dvorak leans heavily (15%) to the right hand, and it changes so many keys which makes learning it a bit hard. I suggest you try Colemak which leans to the same right hand but only 6%, and you only have to learn half the keys of Dvorak. (QWAHZXCVBM stays the same in Colemak.)

[Bravoecho]

In the colemak vs dvorak discussion one point often made is that dvorak is great for writing walls of text, but colemak is more useful on a computer because the common copy/cut/paste keys are NOT moved from the left hand.  So while you are mousing, you can still do those with just one hand.  You can't do that on dvorak.  It may be worth checking workman too, I don't know anything about that layout.

[ksweber]

I think even if most newer keyboard layouts are better than QWERTY they are still very different. If we take for instance only some parameters you already see big differences:

Layout	hand alternation	same finger movements	left hand	right hand	in/out roll	home row	adjacent outer finger pairs	adjacent middle/index finger
QWERTY	52.7%	6.8%	56.9%	43.1%	1	31.6%	8%	13.5%
Workman	55%	3%	50.3%	49.7%	1	63.7%	6.3%	12%
Colemak	58%	1.3%	47.2%	52.8%	1	70%	5.6%	12%
Dvorak type 1b	70.6%	2.5%	45.0%	55.0%	1.5	66.6%	4,1%	6,8%
Malt	54.3%	0.6%	49.6%	50.4%	0.9	60.1% (71.7%)	10,8%	9.4%
AdNW	71.3%	1%	47.8%	52.2%	2.1	69.9%	4,3%	3,8%
NoName32_en	71.5%	1%	47.6%	52.4%	2.7	71.6%	4%	3,7%
NoName30_en	71.1%	1%	47.8%	52.2%	2.8	69.5%	4%	3,8%

Today you can easily create new very good layouts with given optimisation criteria. For instance the following two I created with Dvorak criteria in mind with AdNW optimiser for matrix key layout for english language. As you can see the both rate better than the original of Dvorak. I have one with 32 keys and one with 30 keys. Since the second one is symetric you can mirror the keys if you like the short cut letters on the other side. The keys labeled with '-' you can fill with whatever seldom character you like.

NoName32_en:   ku.,q gpclxz      hieao dtsnrw        y---- vbfmj

NoName30_en:   ku.,q gpclx   hieao dtsnr   y-jz- vbfmw

NoName30_en mirrored:   xlcpg q,.uk   rnstd oaeih   wmfbv -zj-y

Do anyone want to comment on the both no name layouts?

[jakkdl]

I think the one feature you are missing in your two created layouts vs dvorak is "rolling". (don't know it's prevalence in other layouts). If you look at the right-hand letters "htns" the bi-letter combinations th, sn, st etc are fairly common and can be achieved by rolling the hand (pinky first is way more comfortable than index first). Aside from that the layouts look good.

I've got a keylogger running that's up to 1MB that I'm gonna try and use for moving around some letters, numbers and symbols.

[ksweber]

I think the one feature you are missing in your two created layouts vs dvorak is "rolling". (don't know it's prevalence in other layouts). If you look at the right-hand letters "htns" the bi-letter combinations th, sn, st etc are fairly common and can be achieved by rolling the hand (pinky first is way more comfortable than index first). Aside from that the layouts look good.

Thanks for your feedback. I added numbers on rolling and key strokes with adjacent fingers to the table and also updated the layouts. Seems you looked at my very first post where I had copy and paste errors. Please revisit.

ns is more common than sn. Therefore the s is at an more inner position than the n, in order to support inward rolls. As you can see the proportion of inbound rolls to oubound rolls is about 2.7, which means there are almost 3 times more inbound than outbound rolls.

[Oobly]

The "best" layout is very personal and depends on your preferences (inward rolls, outward rolls, alternation, even hand use or biased towards your stronger hand, etc.) and also your physical keyboard layout ("normal" staggered rows layout, matrix layout, staggered columns, number of layers, whether you use thumb key modifiers, etc.).

I use a version of AdNW (optimised for staggered vertical columns and thumb key modifiers, like ErgoDox, Truly Ergonomic and my DIY custom board) and can really see the improvement over QWERTY, especially in home row use. One thing which tends to irritate me with most alternate layouts, though, is the optimisation for inward rolls. I find inward rolls really awkward, outward rolls are much more natural for me, but there isn't a layout designed for outward rolls. So I go for layouts with more hand and finger alternation instead. I know my preference for outward rolls is an exception, though, as most people seem to prefer inward rolls. Just makes it hard to find a layout that suits me. <sigh>

I tried getting the AdNW optimiser to generate a layout for me, but I couldn't understand the instructions properly and have not yet been able to get it to run with my own optimisation criteria. Well, I'm quite used to my modified layout now, so I'll stick with that.

[ksweber]

I tried getting the AdNW optimiser to generate a layout for me, but I couldn't understand the instructions properly and have not yet been able to get it to run with my own optimisation criteria.

the following const double mult_auswaerts = 0.35 sets the amount of badness to count for outward rolls. Per default it is set to 0.35. If you set it to 0 inward rolls are not longer 'rewarded'.

[jakkdl]

Wow those numbers look good, it's a bit tempting to start using one of those layouts on the spot!

Do you have a link to the program used to generate/test the layouts? I'm also interested in what would happen if you start messing with more punctuation and numbers in there - not just "." and ",".

[ksweber]

Do you have a link to the program used to generate/test the layouts? I'm also interested in what would happen if you start messing with more punctuation and numbers in there - not just "." and ",".

Look at www.adnw.de in Download section for the link to the Optimierer sources.

In order to analyse you call : ./opt -b 0  -2 englisch.txt -r bsptast.txt -g out.ps
In order to search for new layouts use: ./opt -b 0  -2 englisch.txt

There is a mailing list where you can get support arround ADNW at adnw@googlegroups.com

[Oobly]

That's AWESOME info! Thanks, ksweber!   

I really like your Noname_30 layout. My layout turned out something like this:

BUQ.X    PCLMF
HIEAO    DTRNS
KY ', -     JGWVZ

Although I have been considering changing it to this:

BU, .Q   PCLMF
HIEAO   DTRNS
KY '- X   JGWVZ

But if I can get the analyser to work and generate a "better" version, that's what I'll do. The word "quiet" is a bit of a finger twister, and "keyboard" isn't great either.

Thanks for all the help.

[ksweber]

Nice to hear that you like NoName30_en. I wanted to show by a sample for english language that there are differences between the discussed alternatives even within the same language.

In fact english is not my primary interest since most of my writing is in german. The AdNW optimiser supports to search for layouts with more than one language as constraint. In order to search for layouts optimised for english & german you have to call:
./opt -b 0 -2 englisch.txt -2 deutsch.txt

In fact AdNW is optimised for both. Therefore the home row looks slightly different as the NoName samples above, optimised only for english.

[ksweber]

Just corrected a typo in the Malt home row percentage - was 30.1% - to be 60.1%.
If someone counts the thumb 'E' to the home row than the home row percentage has to be increased by 11.6% to 71.7%.

[zig_ziglar]

So much science going on. I love it.

Without going into the analytics, I would personally recommend Colemak over Dvorak for obvious reasons (among those discussed). I'm not saying it's "the best", but it's certainly one of the best and is as close to "off the shelf" as you can get for a non-OS-pre-installed layout.

There are several keyboards with hardware level Colemak configurations via DIP switch (Such as the CODE) and the language installation process (if you have admin privileges) is very simple.

Beyond that, it becomes personal preference, as discussed above.

[davkol]

Without going into the analytics, I would personally recommend Colemak over Dvorak for obvious reasons (among those discussed).

I don't see it. What are the "obvious" reasons? All I see are rolls (personal preference), Tarmak (personal preference), some QWERTY-compatible shortcuts (personal preference, a gimmick IMO), preinstalls (significantly worse than DSK), balanced use of hands (personal preference, depends on use case).

OTOH I'm not aware of any studies with actual typists or significant results in typing competitions, while DSK and DSK-like layouts have had great results in typing championships (hint: F-Keyboard, Barbara Blackburn), not to mention countless studies.

/devilsadvocate

[zig_ziglar]

There aren't significant differences between Dvorak and Colemak, but as mentioned before, if you value Ctrl Z,X,C,V, Colemak is the better choice. Colemak is also based on more modern analytics and had the advantage of being able to learn from Dvorak and improve upon it. I can type pretty fast on both ... and a lazy 120 wpm on Qwerty :/

[davkol]

if you value Ctrl Z,X,C,V, Colemak is the better choice

There are several layouts like Minimak. Those change even fewer keys, especially S is kept in its QWERTY position, while offering a big improvement over QWERTY.

Colemak is also based on more modern analytics and had the advantage of being able to learn from Dvorak and improve upon it.

[citation needed]
I have yet to see anything but Shai's comments on his thought process.

[luisbg]

Colemak looks really promising.

Are there online tools to practice typing with this alternative layouts? Most are targetted towards qwerty.

[Tony]

Colemak looks really promising.

Are there online tools to practice typing with this alternative layouts? Most are targetted towards qwerty.

Google "learn Colemak online" to get all websites you need

[luisbg]

Colemak looks really promising.

Are there online tools to practice typing with this alternative layouts? Most are targetted towards qwerty.

Google "learn Colemak online" to get all websites you need

I saw keybr.com has a setting to switch to Colemak. Very nice

[ksweber]

Are there online tools to practice typing with this alternative layouts?

Yes some allow to redefine the keys like http://www.typingtutor-online.com/EN/Aspx/Start.aspx . Having installed Greasemonkey in Firefox you can use the attached file  so_adnw.user.js (1.42 kB - downloaded 251 times.) to redefine the keys to AdNW.

[ksweber]

One thing which tends to irritate me with most alternate layouts, though, is the optimisation for inward rolls. I find inward rolls really awkward, outward rolls are much more natural for me, but there isn't a layout designed for outward rolls. So I go for layouts with more hand and finger alternation instead. I know my preference for outward rolls is an exception, though, as most people seem to prefer inward rolls. Just makes it hard to find a layout that suits me. <sigh>

Why dont you mirror a symmetrical layout in the middle of each hand set. Then you should have a layout with outward rolls preference. For instance the following 'NoName30_en flipped':

NoName30_en:   ku.,q gpclx   hieao dtsnr   y-jz- vbfmw

NoName30_en flipped:   q,.uy wlcpv   oaeih rnstd   -zj-k xmfbg

Layout	hand alternation	same finger movements	left hand	right hand	in/out roll	home row	adjacent outer finger pairs	adjacent middle/index finger
NoName30_en flipped	71.1%	2.7%	47.8%	52.2%	0.4	69.5%	3.5%	7,7%
NoName30_en	71.1%	1%	47.8%	52.2%	2.8	69.5%	4%	3,8%

If you do this flipp in order to minimise inbound rolls (0.4!) then the same finger movements are trippled and the adjacent key strokes with index/middle fingers are doubled.

If you place your index fingers on r&h instead on i&n you get rid of the trippled same finger movements, but you almost tripple the load on the pinkies.

Not a good idea at all!

[ksweber]

Colemak looks really promising.

I do not find it promising. Since it has relatively small hand alternation ratio it suffers from relatively high amount of key strokes with adjacent fingers. Since movements with adjacent fingers are hard to coordinate and slow they should be prevented.
By some small changes on Colemak the number of adjacent key strokes can be reduced significantly.

Colemak: qwfpg jluy; arstd hneio zxcvb km,./

Colemak ||-patch: qfwpg jlyu; asrtd hnoei zxcvb km,/.

Layout	hand alternation	same finger movements	left hand	right hand	in/out roll	home row	adjacent outer finger pairs	adjacent middle/index finger
Colemak	58%	1.3%	47.2%	52.8%	1	70%	5.6%	12%
Colemak ||-patch	58%	1.4%	47.2%	52.8%	1	70%	4.9%	7.8%

[jacobolus]

Since movements with adjacent fingers are hard to coordinate and slow they should be prevented.

This seems like a very oversimplified measure. Typing two keys with adjacent fingers on the same row is usually pretty fast. Typing keys with adjacent fingers from far-apart rows is relatively slow and difficult. And the direction matters too. I find typing a bottom-row index finger and a top-row middle finger is okay, but the reverse (top row index finger & same hand bottom row middle finger) is truly awful.

I think for doing proper ratings, it would be helpful to have some kind of preference (or speed/accuracy/comfort) score for every ordered pair of keys on the keyboard, or at least ones on the same hand. [This is before deciding on what character to associate with anything.] I’m not sure if it would be possible to come up with a fair numerical score for such pairs, but it should at least be possible to come up with an ordered ranking of preferences (in mathematical terms, a total preorder). Obviously an actual scoring of each pair would be most helpful in constructing an optimal layout (or scoring existing layouts) but I’m not sure it’s possible to make one fairly. Even a ranking should be enough to draw some conclusions about the deficiencies of layouts though.

Obviously everyone is slightly different in which fingers they use to type which keys, which keys they find easier or harder to type, etc. (and it depends on the particular keyboard being used). But if a large number of people’s individual preference scores were aggregated, that could be used to generate layouts which were fairly optimized for the average person and also reasonably comfortable for, say, 90% of users.

It would be a bit of a pain in the ass to gather comprehensively, for each person, because even just on each hand separately there are something like 400–500 pairs to consider. Better might be to develop a program which showed two pairs of key-pairs and asked the user to pick a favorite, Hot-or-Not style. But making such data useful would require a large number of participants each adding a substantial number of data points.

[ksweber]

Since movements with adjacent fingers are hard to coordinate and slow they should be prevented.

This seems like a very oversimplified measure. Typing two keys with adjacent fingers on the same row is usually pretty fast. Typing keys with adjacent fingers from far-apart rows is relatively slow and difficult.

Its not fast typing with the pair of ring and middle finger. They are partly using common nervs and tendons. And at least for me making drum-rolls with ring an middle finger is slow motion compared to the pair of index and middle finger. Drum-rolls with the both index fingers are really fast compared. Therefore I try to avoid key strokes with adjacent fingers and favour hand alternation. If someone likes typing with adjacent fingers the Malt layout should be your choice. Here you can compare images showing arcs between keys if bigrams are typed one handed. The more often a bigram occurs the thicker the line is. Direction is going from light to saturated arc end. Here you can see very good the involvement of outer fingers on adjacent key strokes. Vertical arcs are showing same finger movements.

Malt	Colemak

The AdNW optimiser avoids having to type adjacent keystrokes or same finger key strokes on far apart rows. The layouts found usually have very few or no key strokes on far apart rows on the same hand.

I like the idea of searching for nice movements.

[jacobolus]

Its not fast typing with the pair of ring and middle finger. They are partly using common nervs and tendons. And at least for me making drum-rolls with ring an middle finger is slow motion compared to the pair of index and middle finger.

Drum roll meaning alternating back and forth repeatedly? That’s a very different story than just typing one and then the other, once. If you go back and forth a bunch of times in a row, then you're going to need to be relaxing the flexor for one finger (and/or contracting the extensor) at the same time as you’re contracting the flexor for the other finger, and then vice versa. If you just type one and then the other, you can use any common flexors for pressing both, because there’s no special need to release one key before pressing the other [assuming your keyboard can handle 2-key rollover].

Still, you’re right that the index finger is more independently mobile than the middle/ring fingers. Also, I wouldn’t be too surprised if typing pairs of adjacent ring/middle finger keys could lead to more errors from the letters being reversed than other types of key pairs. (I don’t have any data on this point one way or another; I also wouldn’t be too surprised if that’s not true.)

[I’m not an expert, but I bet if you look at professionals’ piano fingerings, you’ll see plenty of 2-3-4-5 or 2-3-4-2 or 2-3-4-1 type movements, but not many like 3-4-3-4-3-4]

[Laiin]

Thinking about switching to colemak when i get an ANSI keyboard (have heard bad things about ISO with colemak/dvorak), but have been wondering about keycap heights and so on, will the keycap heights interfere with the layout? I guess the problem could be solved with blank keys but i dont know about switching layout and using blanks. Also does topre also have different heights on different rows?

First post here on GH, please be gentle  )

[davkol]

have heard bad things about ISO with colemak/dvorak

Huh? Quite the opposite, it enables proper¹ angle/wide mods (like DreymaR's.

will the keycap heights interfere with the layout

Huh²? If you swap around keycaps that aren't uniform profile, it'll be an issue, but why would you do that? O_o

Welcome to GH anyway. ^_~

¹ Still ugly hacks IMHO.

[Laiin]

Huh? Quite the opposite, it enables proper¹ angle/wide mods (like DreymaR's.

Well thats something new (have only seen people with nordic layout talk bad abuot it).

Huh²? If you swap around keycaps that aren't uniform profile, it'll be an issue, but why would you do that? O_o

Well i have a Quick Fire Rapid and i dont think the keys are uniform profile(?), and well if i need to learn colemak i would have to change them or id get pretty confused, and as far as i can tell getting colemak legends(?) on any board is hard unless its The Code Keyboard^[TM] or am i totally mistaken there? 

Welcome to GH anyway. ^_~

Thank you! 

[davkol]

The only common uniform-profile keycaps (for MX stems) are SP's DSA.

But I definitely don't recommend swapping caps around, because it kinda defeats the point of touch typing, doesn't it?

[jakkdl]

Printing a cheat-sheet of the layout and having it besides the keyboard is way superior (imo, duh) to having legends - cause who wouldn't want to learn to touch-type. If you don't you'd need to carry hardware with y ou anywhere if you want to use an alternative layout, touch-typing you only need to change in software.

[jacobolus]

Huh? Quite the opposite, [ISO layout] enables proper¹ angle/wide mods (like DreymaR's.

Huh, that’s quite interesting.

I’ve wondered for a while if it would be possible to:

1. Make an analysis of the most comfortable range of motion for each finger relative to the starting place for the wrist, assuming a 2-d keyboard layout. This would basically result in some kind of score for how hard it is to press a key with any particular finger at any arbitrary position relative to its starting location.
2. Figure out a reasonable range of comfortable hand angles and separation.
3. Take the Apple Japanese laptop keyboard layout, completely ignoring the default character map, and try overlaying the analysis from (1) and (2) over the keys, finding where to put the initial position of each hand such that as many keys as possible were as comfortable to type as possible.
4. Fix those starting positions, and assign each key on the keyboard a finger; so now every key would have a finger and some kind of score for how easy it is to type.
5. Make an optimized layout based on those.

I’m guessing that the home position would be something (referring to physical key positions by their QWERTY names) along the lines of F-E-W-Q for the left hand, and K-O-P-] for the right hand, but the starting position might not be leave fingers precisely resting on those key centers. Then there’d be plenty of useful thumb keys on a Japanese layout keyboard, even easier to hit than normal because of the higher vertical position of the home position.

For a completely custom-made keyboard, it’s nice to be able to place the keys in any arbitrary position/orientation in 3-dimensional space, and include as many or few of them as is practical, add whatever hand split and tenting makes the keyboard most comfortable, etc. But this is less practical for a laptop, which already has a fixed keyboard built-in. [And sometimes I don’t want to carry an external keyboard around.]

[ksweber]

I’ve wondered for a while if it would be possible to:

1. Make an analysis of the most comfortable range of motion for each finger relative to the starting place for the wrist, assuming a 2-d keyboard layout. This would basically result in some kind of score for how hard it is to press a key with any particular finger at any arbitrary position relative to its starting location.

Assuming that nice movements are faster than slow ones, one could get a key logger output to analyse the movement timings from key to key. In order to verify that this is layout independant the data should be gathered from different persons writing comfortable with different layouts.

[jacobolus]

I’ve wondered for a while if it would be possible to:

1. Make an analysis of the most comfortable range of motion for each finger relative to the starting place for the wrist, assuming a 2-d keyboard layout. This would basically result in some kind of score for how hard it is to press a key with any particular finger at any arbitrary position relative to its starting location.

Assuming that nice movements are faster than slow ones, one could get a key logger output to analyse the movement timings from key to key. In order to verify that this is layout independant the data should be gathered from different persons writing comfortable with different layouts.

Well, what I mean is, ignore the existing specific places of any keys. Instead, if we have an assumption of a flat keyboard and an initial hand position, and no other assumptions, then what’s the comfortable movement range of each finger, relative to that initial hand position. So imagine all possible key positions, millimeter by millimeter, and figure out how easy it is for the finger to press them. Or in other words, for each finger, there’s some kind of mapping of two-dimensional position to ease of typing.

Now overlap those functions on an existing keyboard, with the physical key positions already fixed by historical precedent. The goal is to figure out which hand position and orientation leads to the best overall ability to type several keys with each finger in an efficient and comfortable way, throwing out any preconceptions about the proper key layout, the proper way to hold the hands, or the proper assignment of keys to fingers.

[Laiin]

I definitely don't recommend swapping caps around, because it kinda defeats the point of touch typing, doesn't it?

Yeah ofc it does, i just figured it would be easier to change them so if i did go all retard i could just look down.

[zig_ziglar]

The only common uniform-profile keycaps (for MX stems) are SP's DSA.

But I definitely don't recommend swapping caps around, because it kinda defeats the point of touch typing, doesn't it?

Or a IBM model M where the caps can be relocated because the concave profile is generated by the switch mounting plate, not by row-unique cap angles.

[jmolino]

http://www.minimak.org/

I am extremely interested in "Minimak" - it features a graduated learning transition and doesn't stray far from QWERTY.  with only a 4 key change you can realize 60% of the benefit of DVORAK.  With the 12 key change you get ~ 85% of DVORAK.

There is some typing lists available for the typing software trainer "TIPP10" that can help you learn it.
You can easily implement this in software in Linux/Windows/Mac with the files provided in the "Downloads" section on the website.

I've yet to try it as I am trying to first get to 100WPM in QWERTY (I train about 1 hour per day... I'm at ~80WPM now)

Anyone else have any experience w/ MINIMAK?

[Tony]

Thinking about switching to colemak when i get an ANSI keyboard (have heard bad things about ISO with colemak/dvorak), but have been wondering about keycap heights and so on, will the keycap heights interfere with the layout? I guess the problem could be solved with blank keys but i dont know about switching layout and using blanks. Also does topre also have different heights on different rows?

I am a Colemak user and I don't switch any keycaps at all, even if I can. It helps me to get rid of the bad habit of looking at the keyboard while typing.

For reference you can print out the layout in a paper and a wallpaper on your screen. After a week you don't have to look at it anymore.

[jmolino]

I am a Colemak user and I don't switch any keycaps at all, even if I can. It helps me to get rid of the bad habit of looking at the keyboard while typing.
For reference you can print out the layout in a paper and a wallpaper on your screen. After a week you don't have to look at it anymore.

Tony, did you happen to take a look at Minimak?  How do you like this in comparison to Colemak?

I'm debating whether to learn Colemak, or Minimak (probably the 6 key version). 

I want to be able to still retain my QWERTY proficiency but have a better layout available for the times I am able to use it.

[Tony]

Tony, did you happen to take a look at Minimak?  How do you like this in comparison to Colemak?

I'm debating whether to learn Colemak, or Minimak (probably the 6 key version). 

I want to be able to still retain my QWERTY proficiency but have a better layout available for the times I am able to use it.

People love to debate and to wait for the best. I think you guys can wait for another century or two and still there is no best layout at all.

I have switched to Colemak so I don't look at other layouts. I have paid the price; now I don't need to switch to any other layouts again (the switching time is really tough, takes at least 6 months to be fluent.).

For people who are at crossroads, I understand their attachments to Qwerty, since their fear of losing credibility is so imminent. Minimak is good at releasing that fear by give up 20% of efficiency that normally Colemak/Dvorak layout gives, just to make the switching roads easier.

But right now once I have switched, I see that fear has no establishments at all, that's purely psychological. Like that kind of fear you've experienced when you learn to walk or to swim. But you only know it when you have switched, not before.

So I let you guys debate forever, since it's impossible to me to know which layout is better than the other.

But I have converted to Colemak and I decide to call it done.

p/s: I dropped Qwerty completely when I had reached 30wpm in Colemak . If you use two layouts, they share the same motor memory in your brain and the spinal cords, so the confusion is high.

[Ally]

Great! Thanks for sharing

[ideus]

http://www.minimak.org/

I am extremely interested in "Minimak" - it features a graduated learning transition and doesn't stray far from QWERTY.  with only a 4 key change you can realize 60% of the benefit of DVORAK.  With the 12 key change you get ~ 85% of DVORAK.

There is some typing lists available for the typing software trainer "TIPP10" that can help you learn it.
You can easily implement this in software in Linux/Windows/Mac with the files provided in the "Downloads" section on the website.

I've yet to try it as I am trying to first get to 100WPM in QWERTY (I train about 1 hour per day... I'm at ~80WPM now)

Anyone else have any experience w/ MINIMAK?

I will give minimak a try, I found it to be a very interesting idea.