• You are not logged in.

    question for Shai

    • Started by dvomak
    • 3 Replies:
    • Reputation: 0
    • Registered: 01-Oct-2007
    • Posts: 4
    02-Oct-2007 18:39:31#1 question for Shai

    your site states,

    "Same finger typing
    On Dvorak it happens 60% more, and on QWERTY 340% more than Colemak. e.g. "ceded" on QWERTY."

    I have done a number of tests and each time find that Colemak is only about 1-2% more efficient in this
    area.  Such as, when Dvorak has 2.3% same finger, Colemak has .8%, not much of a difference, and
    certainly no way near 60%.  Perhaps I am missing something, but all my tests have been done on your
    site and andrei(sp?)'s. 

    How did you come up with 60%?

    By the way, I do not mean to come off as if I am putting your layout.  I agree with a number of things you have done
    that are improvements over Dvorak, but I have also yet to come up with a text comparison that shows
    that Colemak is superior to Dvorak overall.  Every time Dvorak wins out, but I suppose it comes down to what
    areas one considers important to test in, and what areas are not important.  I think what keeps Dvorak
    with an edge is that Colemak mixes the home row vowels and consonants, and leaves the C where it is
    on Qwerty for the sake of the cut function.  In all the tests I have run, the C is actually a strong
    character, and the M is fairly strong as well.  I totally agree with your home row, the only alternative I
    would consider is keeping AEIOU on home and switching the N with the R, but the boost in home row words without
    the U is very appealing.

    I suppose because I am left-handed in a right-hand dominated world, the splitting up of the cut copy paste
    functions does not bother me.  But beyond this, cut copy and paste are not operations that I ever perform
    during general typing, so to keep the C buried on the bottom row, and the X away from the center of the
    bottom does not make much sense to me, simply for the sake of keeping it easy to do something when i am not typing,
    or in-between typing.  I am sure you have mulled over all this long before you arrived at the final
    layout.  Apart from these things, I think your layout is great, and I am glad to see someone taking a step to improve upon Dvorak
    which can certainly be improved upon.

    So, well done.

    Last edited by dvomak (02-Oct-2007 18:56:03)
    Offline
    • 0
    • Reputation: 2
    • From: Houston, Texas
    • Registered: 03-Jan-2007
    • Posts: 358
    02-Oct-2007 20:09:42#2 Re: question for Shai

    >  when Dvorak has 2.3% same finger, Colemak has .8%, not much of a difference


    If the frequency of same finger is 2.3 % for Dvorak and 0.8% for Colemak,  that
    would mean that same finger instances would occur  65 % less often in your example
    using Colemak than it would using Dvorak.

    100 % * ( ( 0.8 % - 2.3 %)  /  2.3%  )  =  100 % * (-1.5 %  / 2.3 %)  =  - 65 %


    Maybe that sheds some light on where that 60 %  number comes from ?


    That is a signicant difference but not nealy as so compared to QWERTY.   
    It gets more understandable if you play with and compare between many
    suggested layouts the various frequencies of a lot of metrics used in layout optimization design. 

    However, my understanding of the argument is that it not about Colemak being a more
    optimal design than Dvorak.  Both Colemak and Dvorak are far more optimal layouts than QWERTY. 
    The point is that in a world dominated by QWERTY, for gains over QWERTY similar to Dvorak,
    Colemak is much easier and faster for a QWERTY touch typist to learn than Dvorak
    (for all the reasons pointed out in FAQ and elsewhere on the forum).  As a bonus, it's much more
    friendly than Dvorak to all the programs out there that have been written assuming a QWERTY layout.

    The FAQ or elsewhere states that the case is not so compelling for a someone who already is an accomplished
    Dvorak touch typists.

    The issue is not Dvorak versus Colemak.

    Last edited by keyboard samurai (02-Oct-2007 20:14:03)

    http://DRCGH.com | http://texasiaido.com
    http://hi-games.net/profile/14

    Offline
    • 0
    • Reputation: 0
    • Registered: 01-Oct-2007
    • Posts: 4
    02-Oct-2007 20:55:52#3 Re: question for Shai

    hey thanks for the response samurai.  Yes, you are correct.  I should have thought a little further than a surface glance at the percentages.
    Of course, unless you compare it with the whole, you don't get a accurate picture.  For example, in a 1,000 character piece of writing as I
    just tested in the applet the following results are thus:

    Colemak = 12.98 Same Finger Keys
    Dvorak = 40.93 Same Finger Keys
    Difference = 27.95 in favor of Colemak

    Colemak = 732.84 Same Hand Keys
    Dvorak = 493.26 Same Hand Keys
    Difference = 239.58 in favor of Dvorak

    The 28 less keys in Same Finger as stated in the FAQ (60% or so)
    does not appear so great when compared with the
    239 more Same Hand that it garners at the same time.

    Because of the vowel consonant mixing on the home row, there is
    a jump increase in # of words typed with one hand, though,
    not huge.

    But, as you say, it is more about compatibility and ease of
    learning than about smoking Dvorak.

    Last edited by dvomak (02-Oct-2007 21:09:31)
    Offline
    • 0
    • Reputation: 0
    • From: Köln, Germany
    • Registered: 01-Apr-2007
    • Posts: 264
    04-Oct-2007 16:40:06#4 Re: question for Shai

    It's alright dvomak, the percentages always catch people...
    I, for instance, always get confused about the difference between percent and percentage points.
    Or as my father famously said: "Look how the customers are fooled: it says 50% discount, but in fact you only pay a quarter less!" I never really got that, but then again, I didn't want to ask...

    Offline
    • 0