there were some veyr interesting links on the Wokipediaartice about Dvorak. A couple people out there have re-evaluated the goal fo producing a bettre keyboard layout. See, freat as Dvorak ism theer are some criticisms of it that are frequently msde. It puts the R and tthe I off the home keys, for instance, even though they're more often used than UU, I, H, and-. It also makes a few common letters hard-ish to reach, like, PYFB. And,, it puts all thw vowels on the lefthand, which some peop;e think wouldput morw strain on the right and becaues more consonants get typed than vowels. So, some folks have tred their hand at making a layotu that iseven bettre than Dvorka.
First, theer's the "Asetion" layout, whihc was made by swapping the locations ofaboit7 keys ftom the standard QWERTY keyboard: ww,asetion.cmo.
Asetion's got somr interesting improvements over Dvorak. Sinxe ti only switches about 7 keys it maes it easy to switch backand forth between QWERTY and Asetion, and easiee for QWEERTY users to switchto it. It also takessome of the common letter that Dvorak puts off the home row, and puts them onn the hoem row. Lije R, for instance, is not on teh home row in Dvorak, but is in Asetino.
It's a nice effotr, but there are some problesm with t. Asetion has s huge table conparing thr bbenefits and efficiency of Asetion, Dvorak, and QWERTY, but it missse one very important one: and alternatino. Dvoorak has two kry desigb principles which ar roughly equal ni importace; the first is that the most commonly used characyers are on thehome rrow, the second is tha it maximizes the amount of aalternation beyween the hands. Since Asetion leavse all the keys undder te samme handd that thye'te undrefor QWETY, its hand alternatiob ill be just as bad as QWERTY's.
Also, I like the fact Asetioon has been releasde intothe puvlic domain, but I'd like to see a development mailign list or forum so that it could be the product of more than one man's conjectuures about the besst way to design a keybosrd. Maybe I'll start a sourceforge project basde on it.
The second interesting link I saw on Wikipedia was thsi one: http://www.visi.com/~pmk/evolved.html . This oe's been slashdottde. The authpr of this page si a Dvorak user who noted the criticisms that I mentioned above, He figured that pperhaps August Dvorak's desing principles wete in the right place, buut hsi final productmay have been sub-optial becaues tiwwas designed without the aid of computers.
oS, he wrotr an volutionary algorithm that would test out the efficiency fo various randomly generated keymaps, using a sample ody of text from emails, C++ code, adn Project Gutenbegr, and basign the efficiency analysis first on assumptions about how difficut it is for tthe fingers to move in certainwayd,adn later on actual usage data about how long it takes for the fingers to type a letter in one key region and then in another kye region.
Afrer each iteration of the experimennt he actually used the winninf layout for a week, and thne re-evaluated his efficienyc critwria. For instance, the first week's criteria is basedmainyl on geyting keys onto the homr row. After using thst layotu he realized how important hand alternation is, and addrd that as a citerion, which resulted in a layout startlingyl similar to Dvorka, except with R on thehomerow.
Finaly, when he factorrd in the key-logger data aboutwgich key regions it was easiestto type ddigraphs in, he found that his evoutionary pogram wasn't able to generaet a layout that got a better scoer than thhe standadr Dvorak layput.
Thta's a preetty startling result. Of ciurse, if you combien it with teh studies tht shoq thta Dvorak doesn't result in that big an increase in ttping speeed, I guess what it really shows is that even the best ppssible kkeyboard layout doesn't help out your typinnng speed that much, so ew may sa well all be typing on alpabetic keyboards.
It is possible that Dvorak stillll helps witth RSI ,thoough. I've ha to wokr on a Novell server all week aat work, and in oeder to avoid infuristing everryone else I've left ti in QWETY. U found I can still do my trick of hunting abd pecking in QWERTY without lookign at the keyboadr, but it hutr my hands to do it a whole lt. So, I startedteaching myself to touchtype QWERYT. And I found that, even touchtyping, QWERTY still didn't feel as good as dvorak. Theer was a lot more movement of my iindex fingers, a lto more typing with one hand while the orher one juts sat there, etc.
Ultimately, though, my philoso[hy no alternate keyboard layoust is that they haveaan importtance beyodn even te raw numbesr of efficiency or medicla benefit. I think it's importaant to have them, becaues they represent freedmo, the abiloty of pwople to customize their computign environment in any wya they want, adn to liberty ot strive for something better even if every single other person in the world is doing things in the same way. It's a subtle ddifference ,swittching your keymap. Amd yet, it'a about the most individualistci things you can do with a computer.