Xref: utzoo comp.std.misc:174 comp.windows.misc:1170 comp.misc:6629 comp.periphs:1958
Path: utzoo!utgpu!watmath!att!tut.cis.ohio-state.edu!gem.mps.ohio-state.edu!ginosko!uunet!netxcom!netxdev!ewiles
From: ewiles@netxdev.DHL.COM (Edwin Wiles)
Newsgroups: comp.std.misc,comp.windows.misc,comp.misc,comp.periphs
Subject: Re: Chord Keyboards (was User Interface Standards -- *Keyboards!*)
Summary: How to compare keyboards.
Message-ID: <1308@netxcom.UUCP>
Date: 24 Jul 89 23:34:20 GMT
References: <2808@ndsuvax.UUCP>
Sender: news@netxcom.UUCP
Reply-To: ewiles@netxdev.UUCP (Edwin Wiles)
Followup-To: comp.periphs
Organization: /etc/organization
Lines: 96

[ FOLLOWUPS DIRECTED TO COMP.PERIPHS  This being a discussion of	]
[ peripherals after all......						]

[ "nukim@ndsuvax.UUCP" asks how one would go about comparing different	]
[ keyboard arrangements to see which one is the "fastest".  Here's how	]
[ I would go about doing it, combined with some personal opinions....	]

Since it seems that most of the wasted time would be in:
    A) Moving to the new key, and
    B) Restoring your fingers to the 'home' position;

Do a study which indicates how much time is spent (on average) striking keys
which are not directly under a finger.  (If you wanted to get really picky
about it, biomechanics could determine the absolute maximum finger speed given
the ideal hand typing position and the known abilities of the human body, then
it becomes a matter of physics to determine the travel time for each finger to
each non-home key.)

Determine from the hardware what the MAXIMUM typing speed is, presuming that
you are delayed only by the amount of time that it takes for the hardware to
disengage its keys.

Select a representative piece of text to be entered via keyboard.

From that text, you can now calculate the time required by the hardware for
the actual entry of the characters.  This should be closely equivalent for
all the devices (except old style manual typewriters where the letters are
on individual bars that must strike the paper and return to their rest
positions).

Examine the text to determine how many of it's characters are not directly
under the home finger positions for each of the keyboards.  For each of those
characters, add a certain amount of time for the finger to reach the key, and
then return to it's home position before striking the next.

Given that the more you have to move your fingers, the longer it will take to
type a given document; and assuming equal skill in each of the keyboard forms;
it should be obvious that the fewer movements you have to make, the faster you
will be able to type.

Thus, a Querty comes in last, a Dvorak comes in second, and a Chord keyboard
comes in first.

Querty comes in last because it was specifically designed such that the most
used letters would be the farthest apart in the old style manual typewriters.
This was done so that they didn't stick together so much, as they actually had
in the original design which I understand was ordered alphabetically.
(This design is seriously outmoded with the advent of electronic document
printing, and the 'ball'  or 'daisy-wheel' typewriters of today.  It's kept
around because: few of the existing secretaries want to switch over; the
schools will not start teaching any of the new systems until they're in use in
the marketplace; and the manufacturers won't make many of them until they see a
demand.)

Dvorak comes in second, because the most used characters are placed directly
under the user's fingers.  Thus, there is less movement of the fingers.  Thus,
there is less time wasted in moving the fingers about.
(Unfortunately, this makes assumptions about the frequency of characters, which
does NOT hold true for all languages, or all forms of typing.  Question: What
are the most frequently typed characters when one is programming in the "C"
language?  ...when one is writing in French?  ...etc...  One would have to have
either a different keyboard for every purpose/language to achieve the highest
possible benefits.  The problem with this is what does an American in France do
for a keyboard?  Possible answer: Keycaps which have 'LCD' displays on them that
the user could configure with a personal card which is carried around from
keyboard to keyboard.  Of course, this has problems of its own!  Which were
talked about on the network some time ago.)

A Chord keyboard would be the fastest because the only moving finger is the
thumb, and it only has three places to move to.
(The nice thing about the Chord keyboard is that all the keys are in one
unified place.  No more manufacturers coming out with their own 'Querty'
keyboards where the control, alt, return, and shift keys are of varying sizes
and locations!  Of course, they can still mess you up by changing which
"chords" mean what characters, but I think they'd standardize REAL quick.
Either that, or they don't get to sell much equipment!  Most likely, the chords
would be based closely on the ASCII byte values for the characters.)
[No doubt IBM will come out with something based on EBCIDC and then try to
shove it down everyone else's throats as the "de facto industry standard"...
:-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) ]

Now, with ANY keyboard device there IS a learning curve (remember how long
it took you to learn to type without looking at the keyboard?) which means
that a person who knows the Querty keyboard would initially loose a great
deal of speed switching to Dvorak.  But once they got past that learning
curve, they would be MUCH faster.
[IMHO, it would be easier going from Querty to Chord, than it would be from
Querty to Dvorak, since the Chord keyboard is OBVIOUSLY different from Querty.
There's no unconscious training that expects a certain key to be in a certain
location to overcome; only a new set of reflexes to lay down...]

						Enjoy!
..!hadron\   "Who?... Me?... WHAT opinions?!?" | Edwin Wiles
  ...!sundc\   Schedule: (n.) An ever changing	| NetExpress Comm., Inc.
   ...!pyrdc\			  nightmare.	| 1953 Gallows Rd. Suite 300
    ...!uunet!netxcom!ewiles			| Vienna, VA 22180