Not to be a total plug, but FingerWorks MultiTouch technology offers seamless gestures, pointing, and typing on the same surface. Very similar to this MIT work, except you don't manipulate secondary objects, you manipulate fingers on the surface directly. Different finger combinations can attach to different controls like Zoom, Undo/Redo, etc. We'll soon have an SDK available with which you can directly connect zoom/scale/translate hand motions to your favorite GUI controls.
For those who haven't noticed yet, we've been on the market for two years. Haven't taken over yet, but we've made many a power user happy:)
There are some huge differences between our 'zero-force' typing and a membrane keyboard. Membrane keyboards require you to apply pressure at the center of the key. You have to make sure to press hard to always meet the pressure threshold, and you have to make sure NOT to hit between keys, both of which slow you down to speeds 40wpm.
Our 'zero-force' typing system is MUCH more forgiving. There is no minimum pressure threshold, so you can drop your fingers on the surface as lightly as you like. And when you hit in between keys, we have sophisticated drift tracking algorithms and spelling models that guess your intended key correctly most of the time. So with practice people CAN touch type at a pretty good clip, up to 60-70wpm. To avoid long-term drift, we recommend people practice recentering on the home row key 'braille dots' between phrases. You're right, not everyone seems to be able to learn to get by without tactile key edges, but the majority of our customers do.
We're in the process of setting up a community/forum on the FingerWorks site, which could certainly become an organizing hub for TouchStream User Groups. I'll talk to my colleagues about having a 'TouchStream User Group' section.
I think the 'cut battery life in half' was a bit of an exaggeration, unless the reviewer's test was on a mini-notebook. We don't have the final specs yet, but the MacNTouch should use less than half as much power as the TouchStream LP, which uses 250mA@5V -> 1.25 Watts/hour. The 15" PowerBook's battery supposedly has a 61 W-hour capacity. You do the math...
As we're able to reduce the TouchStream LP's power usage, we'll update its tech specs page as well.
This is really a mute point with TouchStreams, since our modifier chords:
http://www.fingerworks.com/modifiers.html
for Shift, Ctrl, and Alt are much nicer than any pinky control key. And if you're an Emacs user, our Emacs mode automatically generates those crazy C-x C-f... sequences from simple gestures.
Basically, TouchStreams have home-row chords and other convenient gestures to replace all these keys. Our customers quickly find the gestures are easier to learn and use than any pinky key placement, hence we focused on large, comfortable placement of the alphabetic keys.
You can remap the BackSpace/Delete or any other key you like with our gesture editor (still in Beta). If you can put up with about a week's confusion from learning any new key swap, the thumb BackSpace will feel great for the rest of your life!
Yes, but in this case you can drop all ten fingers on the surface and rest any time you want. The algorithms see all the fingers come down at once and know you just want to rest.
You DON'T have to move off home row to do the gestures. You can do them and point right on home row. One finger at a time is typing. Two or more fingers are point, drag, scroll, etc. You can Shift by dropping four fingers on home row in the opposite hand. Read the site again:)
The point of the proximity sensing, zero-force capability is that you don't HAVE to tap hard on the surface. You just let gravity drop your fingers on the surface. Imagine drumming your fingers on a tabletop VERY LIGHTLY, so they barely touch. That's all it takes.
Yes, the first day your fingertips might hurt because you're used to tapping so hard on other keyboards, but over a couple days you adapt and type or gesture with minimal motions, like typing with graceful little cat feet:)
If the sensing wasn't 'zero-force', i.e. if it required a minimal mechanical pressure, no matter how small, you would have the problems cited above,
because you would need to apply several times the threshold pressure to ensure the key gets activated. Most people use three or four times the actual spring force to depress mechanical keys, just to make sure the key gets activated each time!
MultiTouch also feels very different from a tiny touchpad. The surface is much larger, so the motion sensitivity doesn't have to be set so high--in fact the motion sensitivity for pointing is about the same as a mouse. Also, instead of those annoying single-finger tap-drags on touchpads, dragging is simply done with three fingers, and scrolling with four. MultiTouch also eliminates double-clicking. A single three-finger tap emits double-click events.
Hello slashdot,
My dissertation, "Hand tracking, finger identification, and chordic manipulation on a multi-touch surface", by Wayne Westerman, is available from www.UMI.com Digital Dissertations as a PDF download:
http://wwwlib.umi.com/dissertations/dlnow/3017731
or you can order a paper copy from them.
I have a few comments on typing and potential health benefits below, but the biggest benefit for most people will be the productivity boost from gestures, which are available on ALL our products.
All MultiTouch gestures can be performed anywhere
on the surface, wherever your hand happens to be. This avoids
wasteful hand repositioning to awkwardly reach for multiple keys
in a hotkey sequence, or to move the mouse cursor onto a toolbar
button.
MultiTouch gestures are distinguished by finger
combination and a simple motion direction as the gesture starts,
so the command assigned to the gesture issues immediately.
In contrast, symbolic gestures with a tablet or mouse cannot be recognized until the symbol is fully drawn.
Drawing complex symbols takes awhile, which delays command issuance and slows you down.
Multi-touch gestures put all common commands IN your fingertips, at will. You can string commands together very quickly and unconsciously, the way the craftsmen picks up and uses tools unconsciously, while his mind remains focused on the workpiece.
We have not completed a formal scientific study of the health benefits. The person who said "there is no evidence that force causes RSI" is partially correct, at least for computer-related RSI. However, I can just as well say there is also no evidence that force does NOT cause RSI. Ergonomists simply haven't been able to properly research the question until now--because nobody had built a full-function 'zero-force' keyboard with which comparative experiments could be done.
In my dissertation I cite several studies of RSI in industrial settings, e.g. meat-packers and assembly-line workers. These studies found that cumulative tissue damage is roughly a product of FORCE X REPETITION above a 'cumulative trauma threshold,' which is basically how fast your tissue can heal itself and dissipate toxins. The force contributions in this equation can be either dynamic, from specific motions, or static, from postural loads (i.e. constant bad posture). Frequent rest breaks help the 'trauma' dissipate before reaching the 'cumulative trauma threshold'.
Our capacitive proximity sensing allows us to minimize the force component--we can actually detect fingers a couple millimeters off the surface, but we threshold for the lightest actual contact with the surface, so both you and the device agree on whether you've touched.
So, given the computer use is inherently repetitive, how do use minimize the REPETITION component? The multi-finger gestures attempt to spread the workload across as many different tissues as possible, employing different sets of fingers in a variety of motions as you do different commands. This way you should be less likely to overuse the muscles and tendons associated with a particular finger. Contrast this to mechanical mouse button clicking, where the same finger has to depress the same button over and over--trauma is more likely to build up in that finger's tendons and muscles, while the other fingers have nothing to do all day! Many ergonomists will at least agree that gripping the mouse too tightly and clicking contribute to RSI--why do you think Microsoft makes the button force so small on their mice? Steve Job's obsession with 1-button mice ensures Mac users can at least spread the clicking force across more fingers, even though we all agree 1-button is a bit confining:)
For surfing and programming, the way the gestures spread workload across the fingers can be very
relaxing. At the end of the day, I feel like I haven't really done any work, UNLESS I got stuck doing some very repetitive dialog operation, having to repeat the exact same gesture hundreds of times in a row. If you're stuck doing something THAT repetitive, it's usually because of bad GUI design--it's time to stop and write a script!
So that's the theory folks. We have many customers who are very happy with it, however, we also acknowledge it does not work for everyone. People with serious finger extensor tendonitis (pain on the TOP of their forearm), for instance, may have trouble with the intermittent hovering needed to type quickly. When typing on our surfaces, you typically drop all ten fingers on home row dimples between phrases to rest and rehome, so fatigue does not build up from hovering. But any hovering at all is painful for people with certain pre-existing injuries).
Again, I don't have statistics on this, but its my impression that zero-force is particularly helpful for customers with moderate finger flexor tendonitis, who may cringe each time they must strike a key. That's what I had and what prompted Elias and I to develop zero-force typing. Before we got the first prototypes working, I could only stand to type about a page per day on a Kinesis Classic, else pain would build up and I would have to rest a few days. After a few months of zero-force typing, I was pain free, and have been able to go back to those 12-hour programming sessions necessary to get product out:)
I have reviewed the epidemiological and ergonomic literature on RSI and ergonomic keyboards as background for my dissertation. I also visited Kinesis Corp. (the president Will Hargraeves is a pretty nice guy) a few years ago and have struggled with forearm tendonitis throughout grad school, so I thought you all might like a summary of my experiences in this area. Much of this info with full references will eventually appear in the appendix of my dissertation:
Split keyboards like the Microsoft Natural effectively address wrist fatigue brought on by extreme wrist postures. When trained typists hold their hands on home row of a conventional keyboard, they force their wrists into "ulnar deviation" (rotation of the wrist outward to get the fingers to lie straight along home row) and "forearm pronation" (twisting of the forearms to get hands to lie on a flat horizontal plane, as opposed to the more neutral forearm rotation when holding a glass). Interestingly, all you two-finger, hunt and peck typists out there may not have been taught to conform your hands to home row and thus may already avoid these bad postures.
The more radical Kinesis design addresses these posture problems too, the main difference being that the Microsoft Natural tends to make you stick your elbows out "like a chicken" (bad) and the Kinesis tends to force you to keep your hands in a fixed position, which is generally bad even if the fixed position is a relatively neutral one. Though the Kinesis reduces the distance you have to stretch for keys, when using any keyboard you should always reach for distant keys with full arm motions from the shoulder. Planting the wrists on the edge of the desk or even a soft wrist rest and reaching with only the fingers strains finger tendons and causes more pressure and friction in the carpal tunnel.
However, posture isn't everything. Epidemiological studies of repetitive strain injuries in industrial settings suggest that soft tissue injury begins to accumulate when the PRODUCT of forces and repetitions applied through a tendon by a given muscle surpass a long-term damage threshold (whose exact value is not known).
Therefore there are three things you can do to avoid exceeding this tissue damage threshold:
1. Reduce repetition. learn to use hotkeys, macros, etc. to improve your efficicency. If you are a programmer who likes long variable names, try the variable name completion feature of Visual Slickedit (which I suggested that they implement when I could barely type. Its pretty cool when a company actually implements your suggestions when you're to crippled to code it yourself--almost wants to make you pay for their product:)
2. Take rest breaks every few minutes, and take a day or two off every once in awhile. This allows your body to clear the accumulated toxins from overused areas and gives microdamage a chance to heal, getting you away from the long-term injury threshold. Over the years I have become so attuned to my pain that I can recognize a specific deep burning pain (unlike superficial soreness) which tells me if I stay at the computer another day I may cripple myself for weeks, so I find something else to do for a couple days till the pain subsides. I could tell you never to push on past this pain regardless of deadlines, but most people (including me) have to learn this lesson the hard way.
3. Use input devices which minimize activation force. For pointing, touchpads from Cirque or Synaptics (the main OEM for laptops) are good because you can click with zero-force taps on the surface. I also have a large mouse-trak trackball which I retrofitted with optical proximity sensors to replace the mechanical buttons. I just place my fingers over the infrared sensors without having to push down and am thereby able to use it 2-3 times as long without pain buildup. Avoid thumb-operated trackballs at all cost--heavy use of them causes a nasty thing called DeQuervain's syndrome. If you insist on a mouse, the Microsoft mouse actually has good easy to press buttons, but they screwed up their keyboard in this respect.
Several studies such as Michael Gerard's dissertation at the University of Michigan, "Effects of Keyswitch Stiffness, Typing Pace, and Auditory Feedback on Typing Force, Muscle Activity, and Subjective Discomfort", have indicated that keyswitches with low make force (.25to.5 Newtons) and long, springy overtravels (key movement past the electrical make point before bottoming out) require less exertion from the finger flexor muscles. However, if the make force threshold is made too small, the keys will no longer support the resting weight of the hands, which will cause extra exertion of the finger extensor muscles to hold the hands up. As part of my dissertation I am exploring ways to allow even lower force typing with full support of resting hands, but much work remains to be done before this can benefit any of you.
The Kinesis contoured models (not the Maxim) have specially designed low force (.30N) keyswitches with long overtravels, and even though the force threshold is so low, their palm rests support the hand particularly well. However, the Microsoft natural uses cheap, stiff, compressible domes instead of real springs under their keys. According to the guys at Kinesis, Microsoft asked 300 people to try different keyboards with different key stiffnesses to see what most people liked best. The problem with this study was that people tend to superficially prefer input devices which feel the same as what they're used to, so Microsoft effectively measured the average stiffness of all keyboards on the market instead of finding the most ergonomic keyswitch. I have a feeling they just went with the compressible domes because of cost, regardless of their study.
As some of you guys have mentioned, stretching and mild strengthening exercises are also very helpful. But if your RSI is getting serious you have probably lost forearm strength (some people start dropping things) and should avoid heavy weights which can cause further injury. I swam for a couple years till I built back enough strength to begin using 5-15 lb dumbbells.
I would say more about speech recognition software, but I'm still using OS/2. IBM refuses to upgrade their discrete dictation product (Voicetype for OS/2) to the continuous version (Viavoice 98) available for Win95, and Voicetype for OS/2 is so slow and annoying that I try to avoid it. I'm sure you Linux users are in a similar predicament. But hey, I like using my hands:)
For those interested, I wrote an essay last year about speech recognition on alternative operating systems for the IACT web site: http://pages.cthome.net/iact/24x7speech1.html
Also, the book "Repetitive Strain Injury: A Computer User's Guide" by Dr. Emil Pascarelli and Deborah Quilter really helped me cope and recognize my bad habits when I was first struggling with RSI.
Slashdot Mirror
Ya, our first bulk shipment to The BackStore, our UK reseller:
http://www.fingerworks.com/resellers.html#uk
will be going out this week. Just talk to them!
Not to be a total plug, but FingerWorks MultiTouch technology offers seamless gestures, pointing, and typing on the same surface. Very similar to this MIT work, except you don't manipulate secondary objects, you manipulate fingers on the surface directly. Different finger combinations can attach to different controls like Zoom, Undo/Redo, etc. We'll soon have an SDK available with which you can directly connect zoom/scale/translate hand motions to your favorite GUI controls.
:)
For those who haven't noticed yet, we've been on the market for two years. Haven't taken over yet, but we've made many a power user happy
http://forums.fingerworks.com
There are some huge differences between our 'zero-force' typing and a membrane keyboard. Membrane keyboards require you to apply pressure at the center of the key. You have to make sure to press hard to always meet the pressure threshold, and you have to make sure NOT to hit between keys, both of which slow you down to speeds 40wpm.
Our 'zero-force' typing system is MUCH more forgiving. There is no minimum pressure threshold, so you can drop your fingers on the surface as lightly as you like. And when you hit in between keys, we have sophisticated drift tracking algorithms and spelling models that guess your intended key correctly most of the time. So with practice people CAN touch type at a pretty good clip, up to 60-70wpm. To avoid long-term drift, we recommend people practice recentering on the home row key 'braille dots' between phrases. You're right, not everyone seems to be able to learn to get by without tactile key edges, but the majority of our customers do.
We're in the process of setting up a community/forum on the FingerWorks site, which could certainly become an organizing hub for TouchStream User Groups. I'll talk to my colleagues about having a 'TouchStream User Group' section.
I think the 'cut battery life in half' was a bit of an exaggeration, unless the reviewer's test was on a mini-notebook. We don't have the final specs yet, but the MacNTouch should use less than half as much power as the TouchStream LP, which uses 250mA@5V -> 1.25 Watts/hour. The 15" PowerBook's battery supposedly has a 61 W-hour capacity. You do the math...
As we're able to reduce the TouchStream LP's power usage, we'll update its tech specs page as well.
And pageup/dn gestures as well! Just slide 4 fingers up/down on the left half. Slide 4 fingers up/down on the right half for smooth scroll.
Just slide two fingers anywhere on the right half, right over the keys! To click, tap two fingertips simultaneously anywhere on the right half:
g ui de.html
http://www.fingerworks.com/touchstream_gesture_
This is really a mute point with TouchStreams, since our modifier chords:
... sequences from simple gestures.
http://www.fingerworks.com/modifiers.html
for Shift, Ctrl, and Alt are much nicer than any pinky control key. And if you're an Emacs user, our Emacs mode automatically generates those crazy C-x C-f
We've posted some notes on the layout page to explain why the modifier and enter keys were relegated to the bottom row.
c NTouch_printable.html
href=http://www.fingerworks.com/images/layouts/Ma
Basically, TouchStreams have home-row chords and other convenient gestures to replace all these keys. Our customers quickly find the gestures are easier to learn and use than any pinky key placement, hence we focused on large, comfortable placement of the alphabetic keys.
Ah, but lefties can use a right hand modifier chord with their left-hand pointing/clicking. Avoids all pinky awkwardness.
You can remap the BackSpace/Delete or any other key you like with our gesture editor (still in Beta). If you can put up with about a week's confusion from learning any new key swap, the thumb BackSpace will feel great for the rest of your life!
Yes, but in this case you can drop all ten fingers on the surface and rest any time you want. The algorithms see all the fingers come down at once and know you just want to rest.
You DON'T have to move off home row to do the gestures. You can do them and point right on home row. One finger at a time is typing. Two or more fingers are point, drag, scroll, etc. You can Shift by dropping four fingers on home row in the opposite hand. Read the site again :)
The point of the proximity sensing, zero-force capability is that you don't HAVE to tap hard on the surface. You just let gravity drop your fingers on the surface. Imagine drumming your fingers on a tabletop VERY LIGHTLY, so they barely touch. That's all it takes.
:)
Yes, the first day your fingertips might hurt because you're used to tapping so hard on other keyboards, but over a couple days you adapt and type or gesture with minimal motions, like typing with graceful little cat feet
If the sensing wasn't 'zero-force', i.e. if it required a minimal mechanical pressure, no matter how small, you would have the problems cited above, because you would need to apply several times the threshold pressure to ensure the key gets activated. Most people use three or four times the actual spring force to depress mechanical keys, just to make sure the key gets activated each time!
MultiTouch also feels very different from a tiny touchpad. The surface is much larger, so the motion sensitivity doesn't have to be set so high--in fact the motion sensitivity for pointing is about the same as a mouse. Also, instead of those annoying single-finger tap-drags on touchpads, dragging is simply done with three fingers, and scrolling with four. MultiTouch also eliminates double-clicking. A single three-finger tap emits double-click events.
My dissertation, "Hand tracking, finger identification, and chordic manipulation on a multi-touch surface", by Wayne Westerman, is available from www.UMI.com Digital Dissertations as a PDF download:
http://wwwlib.umi.com/dissertations/dlnow/3017731
or you can order a paper copy from them.
I have a few comments on typing and potential health benefits below, but the biggest benefit for most people will be the productivity boost from gestures, which are available on ALL our products.
We have not completed a formal scientific study of the health benefits. The person who said "there is no evidence that force causes RSI" is partially correct, at least for computer-related RSI. However, I can just as well say there is also no evidence that force does NOT cause RSI. Ergonomists simply haven't been able to properly research the question until now--because nobody had built a full-function 'zero-force' keyboard with which comparative experiments could be done.
In my dissertation I cite several studies of RSI in industrial settings, e.g. meat-packers and assembly-line workers. These studies found that cumulative tissue damage is roughly a product of FORCE X REPETITION above a 'cumulative trauma threshold,' which is basically how fast your tissue can heal itself and dissipate toxins. The force contributions in this equation can be either dynamic, from specific motions, or static, from postural loads (i.e. constant bad posture). Frequent rest breaks help the 'trauma' dissipate before reaching the 'cumulative trauma threshold'.
Our capacitive proximity sensing allows us to minimize the force component--we can actually detect fingers a couple millimeters off the surface, but we threshold for the lightest actual contact with the surface, so both you and the device agree on whether you've touched.
So, given the computer use is inherently repetitive, how do use minimize the REPETITION component? The multi-finger gestures attempt to spread the workload across as many different tissues as possible, employing different sets of fingers in a variety of motions as you do different commands. This way you should be less likely to overuse the muscles and tendons associated with a particular finger. Contrast this to mechanical mouse button clicking, where the same finger has to depress the same button over and over--trauma is more likely to build up in that finger's tendons and muscles, while the other fingers have nothing to do all day! Many ergonomists will at least agree that gripping the mouse too tightly and clicking contribute to RSI--why do you think Microsoft makes the button force so small on their mice? Steve Job's obsession with 1-button mice ensures Mac users can at least spread the clicking force across more fingers, even though we all agree 1-button is a bit confining
For surfing and programming, the way the gestures spread workload across the fingers can be very relaxing. At the end of the day, I feel like I haven't really done any work, UNLESS I got stuck doing some very repetitive dialog operation, having to repeat the exact same gesture hundreds of times in a row. If you're stuck doing something THAT repetitive, it's usually because of bad GUI design--it's time to stop and write a script!
So that's the theory folks. We have many customers who are very happy with it, however, we also acknowledge it does not work for everyone. People with serious finger extensor tendonitis (pain on the TOP of their forearm), for instance, may have trouble with the intermittent hovering needed to type quickly. When typing on our surfaces, you typically drop all ten fingers on home row dimples between phrases to rest and rehome, so fatigue does not build up from hovering. But any hovering at all is painful for people with certain pre-existing injuries).
Again, I don't have statistics on this, but its my impression that zero-force is particularly helpful for customers with moderate finger flexor tendonitis, who may cringe each time they must strike a key. That's what I had and what prompted Elias and I to develop zero-force typing. Before we got the first prototypes working, I could only stand to type about a page per day on a Kinesis Classic, else pain would build up and I would have to rest a few days. After a few months of zero-force typing, I was pain free, and have been able to go back to those 12-hour programming sessions necessary to get product out
Wayne Westerman (Mr. Light Touch)
I have reviewed the epidemiological and ergonomic literature on
:)
.5 Newtons) and long, springy overtravels
:)
RSI and ergonomic keyboards as background for my dissertation.
I also visited Kinesis Corp. (the president Will Hargraeves
is a pretty nice guy) a few years ago and have struggled with forearm
tendonitis throughout grad school, so I thought
you all might like a summary of my experiences in this area. Much
of this info with full references will eventually appear in the
appendix of my dissertation:
Split keyboards like the Microsoft Natural effectively address
wrist fatigue brought on by extreme wrist postures. When trained
typists hold their hands on home row of a conventional keyboard,
they force their wrists into
"ulnar deviation" (rotation of the wrist outward to get the
fingers to lie straight along home row) and "forearm pronation"
(twisting of the forearms to get hands to lie on a flat horizontal
plane, as opposed to the more neutral forearm rotation when
holding a glass). Interestingly, all you two-finger, hunt and peck
typists out there may not have been taught to conform your hands to
home row and thus may already avoid these bad postures.
The more radical Kinesis design addresses
these posture problems too, the main difference being that
the Microsoft Natural tends to make you stick your elbows out
"like a chicken" (bad) and the Kinesis tends to force you to keep
your hands in a fixed position, which is generally bad even if
the fixed position is a relatively neutral one. Though the
Kinesis reduces the distance you have to stretch for keys,
when using any keyboard you should always reach for distant
keys with full arm motions from the shoulder. Planting the
wrists on the edge of the desk or even a soft wrist rest
and reaching with only the fingers strains finger tendons and
causes more pressure and friction in the carpal tunnel.
However, posture isn't everything. Epidemiological studies
of repetitive strain injuries in industrial settings suggest
that soft tissue injury begins to accumulate when the
PRODUCT of forces and repetitions applied through a tendon by
a given muscle surpass a long-term damage threshold (whose
exact value is not known).
Therefore there are three things you can do to avoid exceeding this
tissue damage threshold:
1. Reduce repetition. learn to use hotkeys, macros, etc. to improve
your efficicency. If you are a programmer who likes long variable
names, try
the variable name completion feature of Visual Slickedit (which I
suggested that they implement when I could barely type. Its pretty
cool when a company actually implements your suggestions when
you're to crippled to code it yourself--almost
wants to make you pay for their product
2. Take rest breaks every few minutes, and take a day or two off
every once in awhile. This allows your body to clear the
accumulated toxins from overused areas and gives microdamage
a chance to heal, getting you away from the long-term injury
threshold. Over the years I have become so attuned to my pain
that I can recognize a specific deep burning pain (unlike
superficial soreness) which tells
me if I stay at the computer another day I may cripple myself
for weeks, so I find something else to do for a couple days
till the pain subsides. I could tell you never to push on past
this pain regardless of deadlines, but most people (including me)
have to learn this lesson the hard way.
3. Use input devices which minimize activation force. For pointing,
touchpads from Cirque or Synaptics (the main OEM for laptops)
are good because you can click with zero-force taps on the surface.
I also have a large mouse-trak trackball which I
retrofitted with optical proximity sensors to replace the mechanical buttons.
I just place my fingers over the infrared sensors
without having to push down and am thereby able to use it 2-3
times as long without pain buildup. Avoid thumb-operated trackballs
at all cost--heavy use of them causes a nasty thing called DeQuervain's
syndrome. If you insist
on a mouse, the Microsoft mouse actually has good easy to press buttons, but
they screwed up their keyboard in this respect.
Several studies such as Michael Gerard's dissertation at the
University of Michigan, "Effects of Keyswitch Stiffness, Typing
Pace, and Auditory Feedback on Typing Force, Muscle Activity,
and Subjective Discomfort", have indicated that keyswitches
with low make force (.25to
(key movement past the electrical make point before bottoming out)
require less exertion from the finger flexor muscles.
However, if the make force threshold is made too small, the keys
will no longer support the resting weight of the hands, which
will cause extra exertion of the finger extensor muscles to
hold the hands up. As part of my dissertation I am exploring ways
to allow even lower force typing with full support of
resting hands, but much work remains to be done before this
can benefit any of you.
The Kinesis contoured models (not the Maxim) have specially
designed low force (.30N) keyswitches with long overtravels, and
even though the force threshold is so low, their palm rests support
the hand particularly well. However, the Microsoft natural uses
cheap, stiff, compressible domes instead of real springs under their keys.
According to the guys at Kinesis, Microsoft asked 300 people to try
different keyboards with different key stiffnesses to see what most
people liked best. The problem with this study was that people tend
to superficially prefer input devices which feel the same as what
they're used to, so Microsoft effectively measured the average
stiffness of all keyboards on the market instead of finding the
most ergonomic keyswitch. I have a feeling they just went with
the compressible domes because of cost, regardless of their study.
As some of you guys have mentioned, stretching and mild
strengthening exercises are also very helpful. But if your
RSI is getting serious you have probably lost forearm
strength (some people start dropping things) and should avoid
heavy weights which can cause further injury. I swam for a
couple years till I built back enough strength to begin using
5-15 lb dumbbells.
I would say more about speech recognition software, but I'm still
using OS/2. IBM refuses to upgrade their discrete dictation
product (Voicetype for OS/2) to the continuous version (Viavoice 98)
available for Win95,
and Voicetype for OS/2 is so slow and annoying that I try to avoid it.
I'm sure you Linux users are in a similar predicament.
But hey, I like using my hands
For those interested, I wrote an essay last year
about speech recognition on
alternative operating systems for the IACT web site:
http://pages.cthome.net/iact/24x7speech1.html
Also, the book "Repetitive Strain Injury: A Computer User's Guide"
by Dr. Emil Pascarelli and Deborah Quilter really helped me
cope and recognize my bad habits when I was first struggling
with RSI.