Author Topic: DataHand questions - designing a keyboard with a finger trough  (Read 16971 times)

0 Members and 1 Guest are viewing this topic.

Offline wolfv

  • Thread Starter
  • Posts: 269
I am considering designing and building a keyboard with a finger trough.
Basically I am trying to squeeze 4 rows of keys inside of 2 inches (50mm), to minimize reach.
The trough keys would be positioned similar to a DataHand, but without side keys between the fingers.
There would be a row of keys a step above the trough, for modifier keys (Ctrl Alt) and mouse buttons.
Here is a side view:


So here are my questions:
1) Is your finger long enough to reach one row beyond the trough, without moving your hand?

2) Would repeatedly lifting fingers out of the trough and onto the modifier/mouse row hurt your fingers, hand, or wrist?
Fingers would only be lifted out of the trough for modifying keys and mousing, so lifting would be less frequent.
Letters, symbols, and arrows would be entirely inside the trough.

3)On the DataHand, how tall is the north finger key measured from the well key? i.e. how far do you lift your finger to clear the north key?

Thank you for sharing your insights.
« Last Edit: Fri, 11 July 2014, 17:03:49 by wolfv »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #1 on: Wed, 09 July 2014, 19:48:01 »
What type of switches are you using, and how are you handling the physical construction of the keyboard?

If you can pick any arbitrary 3-d position/orientation of your keys, I recommend holding your hand in as relaxed a position as possible, and locating the "home row" keys directly at that spot (this will be different heights for different fingers), then without moving your wrist, try to trace the path that your fingers most easily travel when you swivel them inward/outward. For the further-away (distal) rows, you still want to press the keys down in a direction perpendicular to the line from your elbow to your wrist (you don’t want to press outward the way the datahand encourages, IMO), but locate the keys a bit wherever the Z-height is most natural for the finger to rest when it reaches toward that position: for me, this is about 2–3 mm higher for each row further back. For the row just nearer than the home row, I don’t think you want much height difference, but you might want to tilt the key slightly. For the next row nearer, I think your best bet is to rotate the key about 70–90°, and press it by squeezing toward the body instead of pressing downward. This dramatically reduces the amount of wrist movement you need.

For the thumbs, there are many possible layouts, so what’s best probably depends on what kind of inputs you need to get into the computer. E.g. you might want more modifiers (which get held down) or you might want more keys to press and release (e.g. to input characters), or you might want to stick some analog inputs in there like thumbstick/scroll wheel/trackball/trackpad.
« Last Edit: Wed, 09 July 2014, 19:49:47 by jacobolus »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #2 on: Wed, 09 July 2014, 21:18:47 »
Hi Jacobolus.

The keyboard I am currently using is 3 rows of keys on a flat mount plate, plus some thumb keys.
Similar to Suka's BlueCube, but the thumb keys are different.

(picture from http://deskthority.net/workshop-f7/my-diy-keyboard-collection-or-how-i-became-a-kb-geek-t2534.html)

When people type letters, normally their hand floats and fingers only do part of the reaching.
But I can not "float" my hands.
My hand is stationary on a hand rest, all movement comes from my fingers.
After a month of typing my finger joints became sore.
So I want to make a trough keyboard to reduce the most frequent finger movements.

Also, my thumbs are sore from having too many thumb keys.
So I want an additional finger row to off load my thumbs.
The additional row would be dedicated to modifier keys and mouse keys.
I think it would be OK if the modifier/mouse row is a far reach because they are less frequent than letters and symbols.

I will make a cardboard mock up, but that won't tell me how my fingers feel after prolonged use.
The only way to find out is to build the trough keyboard and use it for a month.
It would save me a lot of time if someone could spot an ergonomic flaw in my design before I start building it.

I will use mircoswitches for the fingernail keys http://mykeyboard.co.uk/microswitches/
and Cherry MX for the other keys.
« Last Edit: Wed, 09 July 2014, 21:21:52 by wolfv »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #3 on: Wed, 09 July 2014, 23:20:28 »
Well, good luck! Sounds like a fun project. Let us know how it goes.

Quote
But I can not "float" my hands.
Why not?

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #4 on: Wed, 09 July 2014, 23:42:00 »
Thanks jacobolus,

I can not float my hands because I am on a supine work station, laying flat on my back, and I looking up at a monitor suspended over my head.
My elbows are resting on elbow pads, and my hands are on hand rests next to my hips.
I tried ordinary split keyboards, but for long typing sessions, the hand movements make my arms and shoulders sore.
If DataHand were still in production I would buy one.

Found this picture of a DataHand.
Looks like if the finger wells are a little closer to the hand, the fingers could reach the row of modifier/mouse keys.

(picture from http://geekhack.org/index.php?topic=41422.msg1062059#msg1062059)


Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #5 on: Thu, 10 July 2014, 00:13:47 »
Aha. Well, I haven’t ever tried lying on my back and typing on a split keyboard with my arms straight... only lying on my back using a laptop propped up by my legs, and usually only for <30 minutes, as one among many positions. So I can’t give you super useful specific advice probably. Are you sure that it’s comfortable to type with your arms straight / no bend in your elbows? It seems to me like that would get uncomfortable. (I’d want my elbows at least a bit tilted. Perhaps if your back is not completely horizontal, but at something like a 20–30° angle, you’d be able to have some bend in the elbow without having your hands too elevated.)

Maybe you can post a picture of the way you orient your hands/arms while working, and how you rest your hands? (Or how you plan to after making this planned keyboard.)

Anyway, my general advice would be to try to place your arms where you find them comfortable, and then from there put your wrists in the most neutral position you can. For me, with my arms straight at my sides lying down, it seems like I want at least 45° tenting, maybe even more like 60–70°, for my wrists to be as relaxed as they can from that position. I don’t understand why you wouldn’t be able to somewhat ‘float’ your hands. Maybe some kind of forearm support could help support your arm without putting stress in your wrist/fingers.

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #6 on: Thu, 10 July 2014, 01:25:45 »
Supine brethren unite. This whole "floating" nonsense is for people yet to put their bodies in a position where a rested palm does not mean a palm bearing the weight of the arm.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #7 on: Thu, 10 July 2014, 01:31:21 »
jacobolus,
My elbows are bent 30 degrees and the keyboard is tented 45 degrees.
The setup is easy on the arms, but the current flat keyboard is hard on the fingers.

I considered making a floating arm and hand rest assembly, but it would be quite complicated getting it tuned and balanced.
And the momentum of the assembly would interfere with quick movements.
I think making a keyboard would be simpler.

So now I am deciding on what side of the finger trough to place the modifier/mouse row; the far side or the near side?

There was a similar topic on where to place a track point for DataHand on http://geekhack.org/index.php?topic=41422.msg1062059
On Reply #116, Turbinia wants a track point farther from hand, but it requires finger wells closer to the hand, curling the fingers a little more than ideal.
 

On Reply #119, gbjk wants a track point closer to hand, but pointing bends the wrist and curls the fingers.
 
« Last Edit: Thu, 10 July 2014, 02:43:14 by wolfv »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #8 on: Thu, 10 July 2014, 01:41:39 »
Supine brethren unite. This whole "floating" nonsense is for people yet to put their bodies in a position where a rested palm does not mean a palm bearing the weight of the arm.

Hi Zekromtor,
Nice to meet another supine user :), there isn't many of us around.
That's a good way of explaining it.  It's one of those things that is hard to understand until you experienced it first hand.
Do you lay in a home made setup or do you have a store bought supine workstation?
« Last Edit: Thu, 10 July 2014, 02:44:27 by wolfv »

Offline hoggy

  • * Moderator
  • Posts: 1502
  • Location: Isle of Man
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #9 on: Thu, 10 July 2014, 02:58:14 »
I've seen let alone used a supine work station.  I think you've got a good idea and it's worth progressing.  Some of us here have found 'change' in itself useful... have you discovered the alphagrip and the twidler yet?  I'm only suggesting these as a stop gap/temporary swap measure.
GH Ergonomic Guide (in progress)
http://geekhack.org/index.php?topic=54680.0

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #10 on: Thu, 10 July 2014, 03:01:22 »
So do you currently use a datahand? (or have you used them in the past?) I still don’t quite understand the context your initial questions are coming from.
« Last Edit: Thu, 10 July 2014, 03:03:31 by jacobolus »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #11 on: Thu, 10 July 2014, 03:27:59 »
Hi hoggy,
I have seen the alphagrip online, but haven't seriously considered it.
I can continue using my current keyboard for now, but I need a better long-term solution.

Jacobolus,
I have never seen a DataHand except on line.
I am referencing DataHand because it has some dimensions that are similar to my design.
And I want to learn as much from DataHand as possible before building my own design.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #12 on: Thu, 10 July 2014, 03:53:56 »
Okay. Well, be warned that many or even most of the people who try a DataHand don’t like it. Personally I’ve only tried one for a few minutes, but I found the required motions very awkward and inefficient. (Though to be fair the standard keyboard is also pretty bad.)

I think you might like a design something along the lines of the (never-more-than-prototype) DataStealth / Flinder keyboards, but split:

http://geekhack.org/index.php?topic=55099.msg1236269#msg1236269

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #13 on: Thu, 10 July 2014, 05:31:58 »
jacobolus,
The hand rest on the DataStealth looks a lot like the one I made for my keyboard.
I would like to see the layout for the DataStealth, it must have three layers like DataHand because there are only 46 keys.

You can see the key-cap profiles in the picture, on the left half of the keyboard.
The bottom row keys look like they are angled to be pressed like a gripper.
I am guessing that the bottom row is designed for finger curl, and is used less often, probably modifier and mouse keys.

The user would have to float their hands to reach the top or bottom row.
Bicep muscles are very strong, so it is easy to momentarily lift hands.
Even with just three rows, users are floating their hands half the time, although they are probably not aware of it.
When a user sits, the elbows float.  If the elbows float, the hands float some of the time.
But laying supine, my elbows are firmly planted on an elbow rests, so lifting my hands does not extend my reach.

I think the DataHand keys on the side of the fingers (East and West keys) would be awkward, or at least take some training.
I think finger nail keys (North keys) would be easy to get used to.
So a finger trough might work well; basically a DataHand without vertical keys between the fingers.
The finger trough has three rows within easy reach without moving the hand.
An additional row would be harder to reach, but that's OK if those keys aren't used as often.

I am still contemplating where to place the modifier/mouse row; the far side or the near side of the trough?

Thank you for the picture of the DataStealth, the most ergonomic keyboard for sitters.
« Last Edit: Thu, 10 July 2014, 05:44:00 by wolfv »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #14 on: Thu, 10 July 2014, 06:08:41 »
I would like to see the layout for the DataStealth, it must have three layers like DataHand because there are only 46 keys.
You don’t really need too many physical keys. A standard keyboard has a ridiculous overabundance. My guess is somewhere around 30 is the minimum required number, but ~45–50 might make it slightly easier to do complex stuff.

Quote
You can see the key-cap profiles in the picture, on the left half of the keyboard.
The bottom row keys look like they are angled to be pressed like a gripper.
I am guessing that the bottom row is designed for finger curl, and is used less often, probably modifier and mouse keys.
That's correct. My guess is that these were used for modifiers. Gripping works pretty well for modifiers, and with one modifier per finger, with symmetry between right and left hands, it’s very easy to make multi-modifier chords. This is one of the best layouts I’ve seen as far as modifiers are concerned.

Quote
The user would have to float their hands to reach the top or bottom row.
Bicep muscles are very strong, so it is easy to momentarily lift hands.
Even with just three rows, users are floating their hands half the time, although they are probably not aware of it.
When a user sits, the elbows float.  If the elbows float, the hands float some of the time.
But laying supine, my elbows are firmly planted on an elbow rests, so lifting my hands does not extend my reach.
I don’t quite understand what you’re getting at. The DataStealth thingy should be pretty usable with a completely stationary wrist (or perhaps very slight movement when pressing modifier keys), and I believe the intent was to let typists rest the sides of their palms and still type comfortably. Reaching to the further-away row doesn’t require any overall hand movement, but only extension of the last two finger joints, then flexing of the joint at the base of the finger to actuate the key.

Quote
I think the DataHand keys on the side of the fingers (East and West keys) would be awkward, or at least take some training.
I think finger nail keys (North keys) would be easy to get used to.
I found all of those uncomfortable, but especially the side directions, because individual human fingers don’t really move side-to-side independently with any strength at all, and the muscles for extending the finger at the last two joints are also quite weak. I suspect DataHand users end up bringing bigger muscles to bear by moving their whole hand/arm slightly to actuate those keys, especially the side-to-side ones. The flexors in the finger are much stronger, so the downward and “squeeze”-direction keys are much easier to press quickly and accurately.

Quote
The finger trough has three rows within easy reach without moving the hand.
An additional row would be harder to reach, but that's OK if those keys aren't used as often.
Your additional row is not going to be hard to reach at all: try it.

I recommend you read some papers/book chapters on hand/wrist/finger physiology if you haven’t already, and then do some extensive thinking/observation/testing based on the movements of your own hands. The thumb in particular is quite a complex joint, and so figuring out the optimal position and orientation of keyswitches to be pressed by the thumb is non-trivial. (For the record, I don’t know of any existing keyboard which I think does a good job of taking advantage of the thumb.)
« Last Edit: Thu, 10 July 2014, 06:14:21 by jacobolus »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #15 on: Thu, 10 July 2014, 12:21:47 »
Quote
The user would have to float their hands to reach the top or bottom row.
Bicep muscles are very strong, so it is easy to momentarily lift hands.
Even with just three rows, users are floating their hands half the time, although they are probably not aware of it.
When a user sits, the elbows float.  If the elbows float, the hands float some of the time.
But laying supine, my elbows are firmly planted on an elbow rests, so lifting my hands does not extend my reach.
I don’t quite understand what you’re getting at. The DataStealth thingy should be pretty usable with a completely stationary wrist (or perhaps very slight movement when pressing modifier keys), and I believe the intent was to let typists rest the sides of their palms and still type comfortably. Reaching to the further-away row doesn’t require any overall hand movement, but only extension of the last two finger joints, then flexing of the joint at the base of the finger to actuate the key.
The statement is based experience with my current keyboard which is similar to the DataStealth:
  • hand rest on sides of palms
  • tented at 45 degrees
  • 3 finger rows (DataStealth has 4 finger rows)
The main difference is that my elbows are stationary on elbow pads, so lifting my hands can not extend my reach.
All flex and extend movement comes from my finger joints.
After a month of typing my finger joints became sore.

When I change to an arrangement where I can float my elbow and hands, the fingers are comfortable but may arms and shoulders become sore from the floating hand movement.

That experience makes me thing that floating hand movement is normal for people that sit and use palm rests, even if they are not aware of their floating hand movement.

Quote
The finger trough has three rows within easy reach without moving the hand.
An additional row would be harder to reach, but that's OK if those keys aren't used as often.
Your additional row is not going to be hard to reach at all: try it.

I recommend you read some papers/book chapters on hand/wrist/finger physiology if you haven’t already, and then do some extensive thinking/observation/testing based on the movements of your own hands. The thumb in particular is quite a complex joint, and so figuring out the optimal position and orientation of keyswitches to be pressed by the thumb is non-trivial. (For the record, I don’t know of any existing keyboard which I think does a good job of taking advantage of the thumb.)
I especially need help figuring out something better for the thumb keys.
Everything I have tried starts out feeling good, but after a few weeks, the thumb joints get sore.
Can you recommend some papers/book chapters on hand/wrist/finger physiology?

Thank you.
« Last Edit: Thu, 10 July 2014, 12:46:42 by wolfv »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #16 on: Thu, 10 July 2014, 14:41:57 »
3 finger rows (DataStealth has 4 finger rows)
Yeah, but note: the distal row on the DataStealth is raised up a few millimeters higher than the home row (the Maltron does something similar). Also, the fingers have quite a bit of column stagger and a bit of height difference between columns so that the home row is positioned directly under the fingers in their neutral resting position. The proximal row (which I suspect was for modifiers) is tilted up at roughly a right angle to the others. As a result, there’s much less reaching required.

Do you have a picture of your current keyboard? I suspect that your “3 rows of keys on a flat mount plate, plus some thumb keys” ends up being quite different from an ergonomic perspective.

Quote
The main difference is that my elbows are stationary on elbow pads, so lifting my hands can not extend my reach.
I wonder if there’s some way you could support some/most of the weight of your arms without constraining their movement so much. Seems like that might also help.

I’d still be curious to see some kind of picture of your overall setup.

What made you switch to the supine setup in the first place? Why did you choose it over e.g. a standing desk? I just only know one guy who ever used a supine kind of arrangement, a startup founder/engineer who had severely injured his back from some combination of stress, playing hockey, and lifting heavy boxes up some stairs, and then had back surgery and couldn’t work from a sitting position while he recovered but still needed to write code.

Quote
I especially need help figuring out something better for the thumb keys.
Everything I have tried starts out feeling good, but after a few weeks, the thumb joints get sore.
Can you recommend some papers/book chapters on hand/wrist/finger physiology?
I’m not sure if there’s any particular source I’d recommend. I just went trawling through google scholar on this topic a few months ago, reading whatever I could find, and then spent a bunch of time trying to move and flex my thumb in various positions/directions, to see which were stronger/weaker/easier/harder. If I come across any resources that seem particularly good, I’ll try to remember to link them back in this thread.

I think the Maltron does a pretty good job with its primary two thumb keys. If you can find a Maltron to try somewhere or borrow, I think you could learn a lot from it. Unfortunately the other thumb keys it has aren’t the most conveniently arranged.

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #17 on: Thu, 10 July 2014, 17:06:17 »
Supine brethren unite. This whole "floating" nonsense is for people yet to put their bodies in a position where a rested palm does not mean a palm bearing the weight of the arm.

Hi Zekromtor,
Nice to meet another supine user :), there isn't many of us around.
That's a good way of explaining it.  It's one of those things that is hard to understand until you experienced it first hand.
Do you lay in a home made setup or do you have a store bought supine workstation?

I made a recliner style bed out of 2x4's that's adjustable, memory foam on top (a bit hot). I'm not fully flat right now, but as I do minor adjustments I'm getting closer and closer.  How about you?

Offline SonOfSonOfSpock

  • Posts: 321
  • Location: Colorado, USA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #18 on: Thu, 10 July 2014, 17:44:16 »
Would something like a Kinesis Advantage split in half work for you? Its keywells sound similar to what you want for a trough.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #19 on: Fri, 11 July 2014, 12:48:11 »
Thank  you all for your responses.  I feel like I am making some progress.

I think the Maltron does a pretty good job with its primary two thumb keys.
jacobolus,
I use supine workstation to avoid back pain.  A standing desk is not effective for me.
Are Maltron's two primary thumb keys different than the DataStealth thumb keys?

Zekromtor,
I started with a wooden supine workstation too; then I made a steel one (pictured below).

SonOfSonOfSpock,
I considered the Kinesis Advantage at one time, but it's curved contour reduces reach only slightly.

Here are some pictures of my current keyboard setup.
Left arm position:




Left hand position:






I am leaning towards a DataStealth inspired layout, but with the top row on fingernails (like North keys on DataHand).

Mockup of DataStealth layout based on this picture:










Thumb positioned on top:


Thumb positioned like a gripper:
« Last Edit: Fri, 11 July 2014, 14:01:03 by wolfv »

Offline RED-404

  • Posts: 179
  • Location: 🌎 🇺🇸 Kansas…
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #20 on: Fri, 11 July 2014, 12:58:12 »
So what you're looking for is a modified version of this? Or am I just miss reading.
http://solidartlabs.com/

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #21 on: Fri, 11 July 2014, 13:24:02 »
So what you're looking for is a modified version of this? Or am I just miss reading.
http://solidartlabs.com/
Show Image

Hi RED-404,
Basically I am trying to squeeze 4 rows of keys inside of 2 inches, to minimize reach.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #22 on: Fri, 11 July 2014, 16:02:15 »
I think that mockup is looking pretty good. I recommend trying what you call the "north" keys both (a) using your fingernail-flick-to-press direction, and (b) in the same orientation as the home row keys. I think if they’re raised up 2-3 mm above the home row keys, both reaching for them and pressing them will be pretty comfortable. I think you could even get away with another row beyond that if you wanted, as long as it is raised again by another 2-3mm.

The Maltron has somewhat similar position for its thumb keys, but they’re angled a bit differently in 3-d.

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #23 on: Fri, 11 July 2014, 16:19:47 »
Awesome setup wolfv. It's good to see I'm not the only one taking things to extremes.

Are you a chronic back pain sufferer or did you make your setup simply to be able to comfortably work for long periods of time? I'm more of the latter. I was starting to develop back pain when working for long periods, but I don't have any injuries or chronic issues. I went supine more as a preventative measure, and it was unquestionably the right thing to do. You're pretty close to flat, but I wonder for chronic pain sufferers if perhaps even inverting to a small degree might help.

Looks like you're a little low on padding, though I can't see for sure. How long are you able to lie there comfortably? I'm constantly tweaking my setup with minor adjustments to angle and offsetting the foam for more/less lumbar padding, etc.

Also really like what you have planned there for your keyboard. I could never part with my real mouse, but I too am currently working on a keyboard design with more curvature than the kinesis (which I currently use, modded to hell of course) to have more power at my finger tips. I'm not a big thumb user though as it stands, so that simplifies things for me as I only really need 2 thumb buttons. That said, I have plans for 6 rows of keys, (5.5 really since the top row is only middle finger usable) where all the keys are easily reached by my luckily-long fingers without palm raising. I'll be posting pics a little later in development.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #24 on: Fri, 11 July 2014, 20:50:02 »
Zekromtor,
Laying down is the only thing that stops my back pain.
The 5 degree incline was for my previous keyboard setup.
If you mean "inverting to a small degree" as the head 5 degrees below the feet, like traction; no, I think level is best.
My next supine workstation will be flat and level.
I can put in 12 hour days on my supine workstation, with breaks.

I tried "firm" polyurethane foams at first, but after a few months they would conform to my body.
Then I tried Therm-a-Rest pads, they work extremely well.
I gave up the Therm-a-Rest because my elbows where hitting the Therm-a-Rest (my elbows are below grade and the Therm-a-Rest is wider than the space between my elbows).
The padding I use now is EVA http://en.wikipedia.org/wiki/Ethylene-Vinyl_Acetate
Anti-Fatigue Foam Mat from Harbor Freight http://www.harborfreight.com/4-piece-anti-fatigue-foam-mat-set-94635.html
cut to size and stacked three high, although one or two high would probably be enough.
EVA is easy to cut with a box cutter; I cut a coccyx hole in the top layer.
I have used the EVA for over three years and am happy with it.
Its starting to squish down in spots, so I will probably replace it every 5 years.

jacobolus,
Is this what you mean by?:
  (a) using your fingernail-flick-to-press direction
  (b) in the same orientation as the home row keys
mockup of nail row:


index finger on row 1:


index finger on row 2:


index finger on row 3:

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #25 on: Fri, 11 July 2014, 21:01:35 »
jacobolus,
Is this what you mean by?:
  (a) using your fingernail-flick-to-press direction
  (b) in the same orientation as the home row keys

What I meant is, go ahead and try a DataHand-style version where you extend your finger to actuate the switch, but I recommend also trying a version that’s a bit more like a standard keyboard, where you actuate the switch by flexing your finger downward, but make sure to position the back row a bit higher than the home row.

Your mockup is kind of interesting too though. Definitely experiment. It doesn’t look super comfortable to me, but do whatever works! :)

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #26 on: Fri, 11 July 2014, 21:03:25 »
I can put in 12 hour days on my supine workstation, with breaks.
Seriously? After about 4–5 hours of solid creative work my brain starts feeling pretty fried, and I know it’s time to go for a walk or sit in the park and read a book or something. 12 working hours in a day is just masochism.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #27 on: Fri, 11 July 2014, 21:33:15 »
I can put in 12 hour days on my supine workstation, with breaks.
Seriously? After about 4–5 hours of solid creative work my brain starts feeling pretty fried, and I know it’s time to go for a walk or sit in the park and read a book or something. 12 working hours in a day is just masochism.
Ha Ha. I meant I can be in my supine workstation 12 hours a day, like some people sit 12 hours per day.
I have substituted almost all sitting with supine workstation.
I work in my supine workstation, but I also read ebooks, watch videos, or talk on the phone.

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #28 on: Sat, 12 July 2014, 01:29:45 »
I'll have to try the harborfreight foam. I'm going to feel pretty stupid if it feels better than my expensive memory foam though. You use a normal pillow or do you use that foam even for your head?

er.. just went to order and it's 10 dollars per 2x2feet, not the full mat like I thought. Not gonna break the bank, but not as cheap as I was thinking :)
« Last Edit: Sat, 12 July 2014, 01:31:48 by Zekromtor »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #29 on: Sat, 12 July 2014, 03:28:04 »
Zekromtor,
The head and elbow pads are blue foam cut from a camping pad.
But I installed that before I discovered EVA foam.
The EVA foam should work just as well for elbows and head.
I must warn you though, the Anti-Fatigue Foam Mat is firmer and harder than foams intended for laying on.
It feels weird at first, but it worked for me.
Also notice that I have no padding under my legs:


I bought mine in store, $10 for four 2x2feet squares, I still have a bunch left over.
« Last Edit: Sat, 12 July 2014, 15:15:36 by wolfv »

Offline wolfv

  • Thread Starter
  • Posts: 269
double-sided row
« Reply #30 on: Sat, 12 July 2014, 15:44:41 »
Double-sided row is an idea so crazy it just might work.
The double-sided row has keys on over and under side.

index finger on over side:


index finger on under side:


The double-sided row concept would be useless to people that can sit and float their elbows fore-and-aft.
But for people with elbows fixed to pads, its a way of reaching five rows.
    The layout could be something like:
    • over row - modifier keys
    • under row on finger nails - BkSp Esc Ent Tab Del ' " _
    • 3 base rows - character rows with 3 layers Alpha, Symbols, and Fn

    I simulated the over reach motion on my current keyboard, it's mostly bending the wrist, so the thumb keys end up being a farther reach when fingers are on the over row.
    It feels comfortable, but comfort for long term use is unknown.
    Do you see any down side to using this double-sided row?
« Last Edit: Sat, 12 July 2014, 15:52:36 by wolfv »

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #31 on: Sat, 12 July 2014, 17:16:26 »
I don't like it. The motion to put your finger on the over side is not only a long one but it is a complex one that I think would get annoying if not painful fast.

Have you had you hand physically on a kinesis before? You said the key well wasn't quite compact enough for you, but I'm curious if you're going off the math instead of the actual test. The five rows or real keys (bottom row being a partial) are quite easy to press without palm motion. I have a hunch you are underestimating how much of a difference the curvature makes. I can even hit the 6th row, the crappy rubber F-keys, but only with the middle finger.

I'm not recommending you get one necessarily, but your goal of hitting 5 rows easily seems like a problem that already has a solution. Remember when you do these mockups with cardboard that if the cardboard surface is representative of the switch plate that the curvature will actually be placing the key tops significantly closer together and thus easier to reach. I know from experience that air-typing tests on a curved keyboard backplate can make it seem like the buttons are rather hard to hit, then you place them on, shrinking the radius of the curve, and all of a sudden it's a piece of cake.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #32 on: Sat, 12 July 2014, 20:08:13 »
Zekromtor,
You're right!  I was assuming that Kinesis cap tops where spaced 3/4 inch.
Do you have a Kinesis on hand?

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #33 on: Sat, 12 July 2014, 21:48:45 »
Yep. I can give you whatever approximate measurements you would like.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #34 on: Sat, 12 July 2014, 22:44:05 »
Zekromtor,
That's great.  I am curious how much the concave shape reduces finger reach.
A caliper with it's small-pointy jaws would accurately measure inside the concave well.
On a QWERTY Kinesis Advantage, what is the straight line distance between center of these keys:
   from X to W
   from \ to 2

Thanks.

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #35 on: Sat, 12 July 2014, 23:28:35 »
w to x = 1.29"
\ to 2 = 2.54"
+ to 5 = 3.65" (note the + key is wider than normal and I still went center to center)

Measurements from the left hand side of http://i.imgur.com/IbUWMAP.jpg

Note I do not actually recommend buying a Kinesis. The mods you'd have to do wouldn't make it that much easier than building your own board if you're already proficient at that. Also I _hate_ the elevation (z axis) changes between the columns. I've mentioned before I sanded tons of material off some of the keys and added material to others just to get them closer to level. Having to hold my pinky up to use my middle finger buttons caused a lot of pain before that issue was fixed. When your fingers are curled over, there is not much difference in elevation, and this keyboard really overcompensates.

With the curvatures of the Kinesis though you are saving over 15% distance on Y axis distance from key to key, and the Kinesis does not curve all that much. You can definitely do a more aggressive curve and have it still be comfortable to type on, but you'll probably only be able to do it in one direction.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #36 on: Sun, 13 July 2014, 00:23:13 »
That's great.  I am curious how much the concave shape reduces finger reach.
A caliper with it's small-pointy jaws would accurately measure inside the concave well.
On a QWERTY Kinesis Advantage, what is the straight line distance between center of these keys:
   from X to W
   from \ to 2
Those aren’t quite the measurements you want. What you want to know is, with the wrist fixed, and the joint at the base of the finger fully extended, as you extend the other two finger joints how far are the key tops from the path the fingertip takes. The straight line chord length between key-tops isn’t super meaningful by itself.

Again, for a bowl-like layout, I think you want something closer to the Maltron (or DataStealth) than the Kinesis... i.e. raise the row up, but still keep the switch pointed along the same axis.
« Last Edit: Sun, 13 July 2014, 00:24:54 by jacobolus »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #37 on: Sun, 13 July 2014, 00:23:56 »
Thanks Zekromtor, much appreciated.
I will shrink the y dimension on my next layout at bit.
I looked for 3/8" x 1/2" key-cap tops (so fingers can feel key centering) but didn't find any.

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #38 on: Sun, 13 July 2014, 00:27:09 »
I looked for 3/8" x 1/2" key-cap tops (so fingers can feel key centering) but didn't find any.
Non-square keycaps don’t really exist as a commercial product today, but Matias is going to produce .75x1 unit keys in their upcoming ErgoPro for the arrow keys if you’re willing to wait a few months, or you could always try to produce your own keycaps. (Kurplop did this: read through http://geekhack.org/index.php?topic=47787.0 for some of the details.)

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #39 on: Sun, 13 July 2014, 00:39:52 »
jacobolus,

I know what you mean.
I have traced my finger tip travel with a clipboard between my fingers.
So I will build some curve into the columns.

For people with elbows fixed to pads, reach in the y direction is really limited.
So I will add a fingernail row with: BkSp Esc Ent Tab Del ' " _

Thank you for your input :thumb:.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #40 on: Sun, 13 July 2014, 01:15:52 »
Thanks jacobolus :).
The mantais .75x1 keycaps look nice.  http://matias.ca/ergopro/pc/
I emailed matias asking where I could buy some of those key caps.

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #41 on: Sun, 13 July 2014, 01:17:25 »
With a combination of curvature and a fingernail row, you'll have all the keys you need I'd think.
I'm having to make my own keycaps as well. If you think finding the right cherry caps is hard, try finding any alps key caps, let alone specialty ones. Matias is going to be the only game in town for that soon.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #42 on: Sun, 13 July 2014, 03:31:36 »
jacobolus,
I guess I am stuck with standard square caps :(.
What do you think of these ALPS caps?
http://www.alibaba.com/product-detail/Keyboard-Switch-Cap-Key-pannel-switch_609204226.html

Yes, more keys and less reach, that's got to be an improvement :).
Now I can design a layout.  Will post it here in a few days.

Thanks for all your help.


Offline wolfv

  • Thread Starter
  • Posts: 269
keyobard layout
« Reply #43 on: Mon, 14 July 2014, 06:37:48 »
The guidance you have provided has already resulted in a better keyboard design.
The DataStealth-inspired bottom-row angle and tighter key spacing in the Y direction came from listening to you.

Here is the layout I plan on using for my next keyboard (spread sheet attached).
The Alt key and Ctrl key are sticky.
I was careful, but there might be some things about keyboard layout designing that I am not aware of.
Please let me know if you notice any flaws in the design.

Thank you.

« Last Edit: Tue, 22 July 2014, 17:45:36 by wolfv »

Offline jacobolus

  • Posts: 3661
  • Location: San Francisco, CA

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #45 on: Mon, 14 July 2014, 16:01:48 »
I'm a proponent for more keys and less layers, but the layouts you've made for your setup don't have any flaws that I can spot. One thing I'm testing out with my new board is having all 10 number keys, the enter key, and delete key all on the left hand. It's not as much an issue for you, given your mouse setup, but for me it is incredibly beneficial when programming to have access to those keys that are used so frequently, specifically when one hand is preoccupied on the mouse.

Sticky Alt and Ctrl makes sense for you given the fingernail activation, but won't that make copy/paste a pain in the ass? Maybe if you programmed a special case scenario where a quick double tap on alt was copy and a quick double tap n ctrl was paste that would save you from having to do the extra keystroke.

And are your layer keys sticky? I've always used temporary activated layers, but that may be what makes me less fond of layers.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #46 on: Tue, 15 July 2014, 03:40:02 »
Sticky Alt and Ctrl makes sense for you given the fingernail activation, but won't that make copy/paste a pain in the ass?
I don't see how sticky ctrl would make copy & paste cumbersome.
Sticky and non-sticky copy & paste are the same because Ctrl and the letters 'C' 'V' are on opposite hands.
Sticky is needed when the letter to be modified is on the same hand as the Ctrl key.
e.g. underline selected text in Word:
   Ctrl+U
where Ctrl and the letter 'U' are both on the right index finger, making simultaneous key press difficult.

Holding left mouse button key with fingernail for a long time while selecting text might be strenuous.

Quote
And are your layer keys sticky? I've always used temporary activated layers, but that may be what makes me less fond of layers.
You're making me rethink layer keys (that's a good thing).
I added an 'Alpha' layer key to the layout.  A spreadsheet of the updated layout is attached.

They are 2-mode-layer keys.  Here is how it works:
There is always exactly one one locked layer, the other layers are unlocked.
The locked layer is active when no layer key is being held down.

When a layer key is pressed, its layer becomes temporally activate.
When the layer key is released,
   if no other key was pressed, the layer becomes locked
   else another key was pressed, and the locked layer becomes active again.

Table of layer usage:
Code: [Select]
LAYERS
 TASK ALPHA SYM Fn UNLOCKED LAYERS TEMPORARILY USED FOR
 typing locked insert symbol, move cursor
 data entry locked insert letter, move cursor
 navigation locked insert letter or symbol

This 2-mode-layer key is something I invented, and it has not been tried as far as I know.
Please let me know if you see a flaw, or know of a similar implementation.

« Last Edit: Tue, 15 July 2014, 03:45:42 by wolfv »

Offline Zekromtor

  • Posts: 241
    • My Setup
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #47 on: Tue, 15 July 2014, 04:47:02 »
Sticky Alt and Ctrl makes sense for you given the fingernail activation, but won't that make copy/paste a pain in the ass?
I don't see how sticky ctrl would make copy & paste cumbersome.

Perhaps cumbersome wasn't the right word. I just meant there's potentially an extra click involved -- not if you're immediately pasting after you copy, CTRL C V CTRL, but if you want to do things in between a copy and paste you have to CTRL C CTRL, *do stuff that requires CTRL not be held* CTRL V CTRL right? Even so though, I did not realize you were using opposite hands for the action, which could be argued is more important than saving a click. I just like the idea of using double taps on keys that would never have a double tap under normal circumstances for special macros, though I'm yet to do it myself. For instance, you might be able to simplify your setup and remove the need for an alpha layer key by making a double tap on any of the other layer keys immediately return you to alpha.

I like your conditionally sticky system for layers. I haven't done anything quite so fancy, but currently my main thumb key is multifunctional - hold it down and it's a temporary layer shift, tap it and it's the usual spacebar. Just another example of exploiting that the spacebar is not a key that you'd ever normally hold down. Not compatible with your system, though.

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #48 on: Tue, 15 July 2014, 05:58:29 »
Zekromtor,
A sticky is not a toggle, which seems to be what your examples are.
StickyKeys allow the user to press and release a modifier key and have it remain active until any other key is pressed.
Immediately pasting after you copy, one finger method:
   CTRL C CTRL V
Sticky keys can also function as normal non-stickies, no extra key press if you can use two fingers.
Immediately pasting after you copy, while holding CTRL:
   CTRL CV
Edit between a copy and paste:
   CTRL C edit... CTRL V

Turn on the Sticky on your operating system and try it.
« Last Edit: Tue, 15 July 2014, 06:19:12 by wolfv »

Offline wolfv

  • Thread Starter
  • Posts: 269
Re: DataHand questions - designing a keyboard with a finger trough
« Reply #49 on: Tue, 15 July 2014, 10:40:59 »
Thanks jacobolus :).
The mantais .75x1 keycaps look nice.  http://matias.ca/ergopro/pc/
I emailed matias asking where I could buy some of those key caps.

Matias Help Desk said:
Quote
Thanks for your email and interest in our products.
In early September we are planning to sell new sets of keycaps on our web site.
There will be four choices:
- White keys blank
- Black keys blank
- White keys printed
- Black keys printed

A complete set is 107 keys x $0.20 each + shipping.
We have not discussed selling smaller quantities yet but I will mention it to our VP.
Regards,
--
Matias Help Desk