5 ms·

Show HN: Engelbart's Chorded Keyboard - Built at SHDH42 yesterday

43 points by ptarjan 16y ago

Yzupnick 16y ago
Why is it lefty?
- jodrellblank 16y ago
  I guess because Englebart's was. http://www.cc.gatech.edu/classes/cs6751_97_fall/projects/follow_me/hw4/first_demo_large.jpg http://www.cc.gatech.edu/classes/cs6751_97_fall/projects/fol... Right hand for navigating (mousing), left hand for typing.
ptarjan 16y ago
I was inspired by Englebert's "Mother of all Demos" shown at Super Happy Dev House 42 and decided to actually try out the device that was in his left hand.
- jf 16y ago
  How did it look on the big screen? We got a DVD copy of the original tapes from the Stanford library!
  - idoh 16y ago
    I was there, it was rad. When it ended there was a some applause from the people in the room.
  - ptarjan 16y ago
    It was amazing on the IMAX. I was a little sad that it only displayed on about 1/5th of the screen infront. I assume it only fit onto one projector instead of the software cutting it up to fill the whole dome. It is a little slow at first, but once he gets to explaining the tech, it was wonderful.
gnosis 16y ago
Cute. But I'd need a lot more than this for programming, which uses many special symbols. Even regular English requires being able to type quotes, question marks, exclamation points, numbers, etc.. Also, as far as the implementation of this particular typing tutor goes, it needs to let the user retry to type a given chords until he gets it right, instead of immediately moving on to the next one. And the chords should be taught in small batches, rather than trying to teach them all at once. While we're on the subject of chording, I should mention that there's a vim plugin that will let you map chorded keystrokes: http://www.vim.org/scripts/script.php?script_id=2425 http://www.vim.org/scripts/script.php?script_id=2425
kadavy 16y ago
This was really interesting to play with. I kept getting caught up cognitively on the fact that the thumb corresponded to the 5th box in all of the chord diagrams. I feel like some visual differentiation would have made that easier to grok.
- rogercosseboom 16y ago
  ^^this I also had the same issue, oddly enough it got worse as i got more accustomed to the program. I think mainly because I was visually counting the boxes from the right to left. maybe instead of boxes use a numbering or finger image? But yeah definitely some differentiation between the 'f' key and the 'space' key
- ptarjan 16y ago
  Great idea. I tried to do a <sub> and it looked funny, and adding a full   was too wide. Maybe I'll play with a smaller space character.
  - agscala 16y ago
    I don't think you can make it too wide. Even if it was twice as wide or three times as wide it would be good enough for the brain to grok
pkamb 16y ago
It's funny how it's not optimized at all. "E" and "T" are both incredibly hard to type; this would give you RSI problems in no time.
- ptarjan 16y ago
  Yeah, exactly. He just enumerated the binary representation. A=1, B=2, C=3... Maybe I should build a Dvorak-ish version of this...
- TeMPOraL 16y ago
  Isn't keyboard-related RSI caused by moving wrists? This system doesn't require you to move your hand at all. BTW. I've just read on Wikipedia that there's no causation proven between RSI/CTS and keyboard typing. Interesting, as from what I've seen, it's a widespread belief.
mbreese 16y ago
I found it very hard to do this with my left hand. Especially since the first bit is so widely used and my pinky on that side is pretty weak. I guess you can tell that I'm not an emacs user :)
- ptarjan 16y ago
  Yeah, totally agree. The key mappings need some work. I just wanted to emulate his original one. I'll do a Dvorak one, might be easier.
makmanalp 16y ago
I can't type G for the life of me. Maybe it has to do with the number of simultaneous keypresses on my keyboard?
- yurylifshits 16y ago
  After 5 minutes: I can not get G either!!! Paul, can you change time intervals to to catch less accurate chors?
  - ptarjan 16y ago
    I have it set to 125ms. I bumped it to 200 and it felt really unresponsive. When in doubt, make a configuration option right? ;)
- agscala 16y ago
  It's a problem with keyboard ghosting. That's the problem with chording keys like this, once you're hitting >= 3 keys, you'll find some just won't work.
wazoox 16y ago
I'm a pianist; it was easy as cake, actually would be even easier with finger numbers, or represented as notes :)
dmcg 16y ago
I spent a happy year with a MicroWriter chorded keyboard in the early 90s. My right hand still knows the letters, so there must be something in it.
- ptarjan 16y ago
  Neat. What was they keyboard layout? The same as this one or was it frequency based?
  - jodrellblank 16y ago
    I believe they are the same as the descendent https://sites.google.com/site/cykeybellaire/cykey-codecard https://sites.google.com/site/cykeybellaire/cykey-codecard
pz 16y ago
is there a standard mapping for chorded keyboards? i feel like X has a much easier combo than E. would be interesting to use data on people's error rate for different key combos and then remap frequently used letters to easier combos.
- gnosis 16y ago
  The only standards I'm aware of are standards for stenotype machines (chorded keyboards that stenographers) use: https://secure.wikimedia.org/wikipedia/en/wiki/Stenotype https://secure.wikimedia.org/wikipedia/en/wiki/Stenotype And there are a fair number of competing standards and keyboard layouts (especially if you count layouts/standards for various languages). Stenotypes machines and musical instruments are really the only chorded keyboards in wide enough use to develop any kind of standard.
reeses 16y ago
I remember reading about Steve Roberts building chording keyboards into his "handlebars" on his Winnebiko years ago. Since he had well more than 2^5 bits using both hands[1], he was able to assign macros that helped significantly with his effective typing speed. Englebart says that his layout is not intended for speed typing, with the typical trained typer being able to type 30-50wpm. He observed that typing speed on conventional keyboards does not have a strong correlation with typing speed on his chording layout. Sight-reading piano players have a bit of an advantage here, of course. I've always wanted one of the Twiddlers, and am kind of excited they're back on the market, but I'd vastly prefer a wireless version. I've also been looking forward to the release of the FrogPad. It would be nice to have a better-than-touchscreen keyboard for emails on the iPad or iPhone. [1] I can't remember the exact layout now. If it used all ten fingers, or if it had multiple positions for, say, thumbs, à la the BAT keyboard, it's gone from my memory. Regardless, it used both hands.
js2 16y ago
If you're looking for a one-handed chorded keyboard on which to type, the BAT has been around for many years: http://www.infogrip.com/product_view.asp?RecordNumber=12 http://www.infogrip.com/product_view.asp?RecordNumber=12
biot 16y ago
Interesting idea, but it's rather odd that the keystroke for P is pressing <SPACE>, but the keystroke for <SPACE> is a four character combination when <SPACE> should be its own chord. While the binary incrementing keystroke combination is somewhat logical, I would imagine that a more frequency-based assignment would ultimately be more successful as muscle memory is far more important than remembering that the chord for V is binary+1 of the chord for U.

43 points by ptarjan 16y ago

Yzupnick 16y ago

Why is it lefty?

jodrellblank 16y ago
I guess because Englebart's was. http://www.cc.gatech.edu/classes/cs6751_97_fall/projects/follow_me/hw4/first_demo_large.jpg http://www.cc.gatech.edu/classes/cs6751_97_fall/projects/fol... Right hand for navigating (mousing), left hand for typing.

ptarjan 16y ago

I was inspired by Englebert's "Mother of all Demos" shown at Super Happy Dev House 42 and decided to actually try out the device that was in his left hand.

jf 16y ago
How did it look on the big screen? We got a DVD copy of the original tapes from the Stanford library!
- idoh 16y ago
  I was there, it was rad. When it ended there was a some applause from the people in the room.
- ptarjan 16y ago
  It was amazing on the IMAX. I was a little sad that it only displayed on about 1/5th of the screen infront. I assume it only fit onto one projector instead of the software cutting it up to fill the whole dome. It is a little slow at first, but once he gets to explaining the tech, it was wonderful.

gnosis 16y ago

Cute. But I'd need a lot more than this for programming, which uses many special symbols. Even regular English requires being able to type quotes, question marks, exclamation points, numbers, etc.. Also, as far as the implementation of this particular typing tutor goes, it needs to let the user retry to type a given chords until he gets it right, instead of immediately moving on to the next one. And the chords should be taught in small batches, rather than trying to teach them all at once. While we're on the subject of chording, I should mention that there's a vim plugin that will let you map chorded keystrokes: http://www.vim.org/scripts/script.php?script_id=2425 http://www.vim.org/scripts/script.php?script_id=2425

kadavy 16y ago

This was really interesting to play with. I kept getting caught up cognitively on the fact that the thumb corresponded to the 5th box in all of the chord diagrams. I feel like some visual differentiation would have made that easier to grok.

rogercosseboom 16y ago
^^this I also had the same issue, oddly enough it got worse as i got more accustomed to the program. I think mainly because I was visually counting the boxes from the right to left. maybe instead of boxes use a numbering or finger image? But yeah definitely some differentiation between the 'f' key and the 'space' key
ptarjan 16y ago
Great idea. I tried to do a <sub> and it looked funny, and adding a full   was too wide. Maybe I'll play with a smaller space character.
- agscala 16y ago
  I don't think you can make it too wide. Even if it was twice as wide or three times as wide it would be good enough for the brain to grok

pkamb 16y ago

It's funny how it's not optimized at all. "E" and "T" are both incredibly hard to type; this would give you RSI problems in no time.

ptarjan 16y ago
Yeah, exactly. He just enumerated the binary representation. A=1, B=2, C=3... Maybe I should build a Dvorak-ish version of this...
TeMPOraL 16y ago
Isn't keyboard-related RSI caused by moving wrists? This system doesn't require you to move your hand at all. BTW. I've just read on Wikipedia that there's no causation proven between RSI/CTS and keyboard typing. Interesting, as from what I've seen, it's a widespread belief.

mbreese 16y ago

I found it very hard to do this with my left hand. Especially since the first bit is so widely used and my pinky on that side is pretty weak. I guess you can tell that I'm not an emacs user :)

ptarjan 16y ago
Yeah, totally agree. The key mappings need some work. I just wanted to emulate his original one. I'll do a Dvorak one, might be easier.

makmanalp 16y ago

I can't type G for the life of me. Maybe it has to do with the number of simultaneous keypresses on my keyboard?

yurylifshits 16y ago
After 5 minutes: I can not get G either!!! Paul, can you change time intervals to to catch less accurate chors?
- ptarjan 16y ago
  I have it set to 125ms. I bumped it to 200 and it felt really unresponsive. When in doubt, make a configuration option right? ;)
agscala 16y ago
It's a problem with keyboard ghosting. That's the problem with chording keys like this, once you're hitting >= 3 keys, you'll find some just won't work.

wazoox 16y ago

I'm a pianist; it was easy as cake, actually would be even easier with finger numbers, or represented as notes :)

dmcg 16y ago

I spent a happy year with a MicroWriter chorded keyboard in the early 90s. My right hand still knows the letters, so there must be something in it.

ptarjan 16y ago
Neat. What was they keyboard layout? The same as this one or was it frequency based?
- jodrellblank 16y ago
  I believe they are the same as the descendent https://sites.google.com/site/cykeybellaire/cykey-codecard https://sites.google.com/site/cykeybellaire/cykey-codecard

pz 16y ago

is there a standard mapping for chorded keyboards? i feel like X has a much easier combo than E. would be interesting to use data on people's error rate for different key combos and then remap frequently used letters to easier combos.

gnosis 16y ago
The only standards I'm aware of are standards for stenotype machines (chorded keyboards that stenographers) use: https://secure.wikimedia.org/wikipedia/en/wiki/Stenotype https://secure.wikimedia.org/wikipedia/en/wiki/Stenotype And there are a fair number of competing standards and keyboard layouts (especially if you count layouts/standards for various languages). Stenotypes machines and musical instruments are really the only chorded keyboards in wide enough use to develop any kind of standard.

reeses 16y ago

I remember reading about Steve Roberts building chording keyboards into his "handlebars" on his Winnebiko years ago. Since he had well more than 2^5 bits using both hands[1], he was able to assign macros that helped significantly with his effective typing speed. Englebart says that his layout is not intended for speed typing, with the typical trained typer being able to type 30-50wpm. He observed that typing speed on conventional keyboards does not have a strong correlation with typing speed on his chording layout. Sight-reading piano players have a bit of an advantage here, of course. I've always wanted one of the Twiddlers, and am kind of excited they're back on the market, but I'd vastly prefer a wireless version. I've also been looking forward to the release of the FrogPad. It would be nice to have a better-than-touchscreen keyboard for emails on the iPad or iPhone. [1] I can't remember the exact layout now. If it used all ten fingers, or if it had multiple positions for, say, thumbs, à la the BAT keyboard, it's gone from my memory. Regardless, it used both hands.

js2 16y ago

If you're looking for a one-handed chorded keyboard on which to type, the BAT has been around for many years: http://www.infogrip.com/product_view.asp?RecordNumber=12 http://www.infogrip.com/product_view.asp?RecordNumber=12

biot 16y ago

Interesting idea, but it's rather odd that the keystroke for P is pressing <SPACE>, but the keystroke for <SPACE> is a four character combination when <SPACE> should be its own chord. While the binary incrementing keystroke combination is somewhat logical, I would imagine that a more frequency-based assignment would ultimately be more successful as muscle memory is far more important than remembering that the chord for V is binary+1 of the chord for U.