I love chord keyboards as a concept, but I've never committed to one because I've been wary of acquiring a skill on such a specific device, and then not having access to one unless I bring it with me. Making a chorded keyboard a software solution is very appealing for that reason, especially if ASETNIOP were to become a standard approach.
For what it's worth, the first mouse was intended to be used in conjunction with a chorded keyboard when Doug Engelbart invented it in 1963: http://dougengelbart.org/images/pix/img0030.jpg
I've tried the online tutorial, and it's surprising how quick I've noticed some muscle memory for the most frequent chords (for letters r, h, d, l).
I think they've found something with the "always use the same finger as in QWERTY, plus another one". I was skeptical that users could ever adapt to a new layout, but this isn't exactly that, so it has potential.
The best thing is that it doesn't interfere with the existing skill on the QWERTY keyboard.
People in this topic might be interested in a project of mine: one-handed typing using your existing two-handed muscle memory.
Basically the muscle memory you use to type "F" is the same as typing "J"; index finger, home row. Your left hand already has the muscle memory to type right-hand characters, and vice versa. Press the "mirrored" key with the other hand... your brain should be able to make this substitution almost automatically.
Add some software predictive text and you can suddenly type fast with one hand with hardly any learning. Especially good for temporary hand injuries.
Sounds like really cool software! Is it agnostic of my keyboard layout? Dvorak user here.
That would mean I can not only freak out people that my keys don't type what they say, but that I only need one hand and somehow it works out what I want. People will surely take me for a con artist.
Is caps-lock toggling the only way to use it in the paid app, or does the menu item turn it on and off? Do numbers and other coding-symbols get chorded in the main app? Also, is there any research into how long it takes to go from Dvorak to this sort of style?
Thanks for trying the app. How does the technique work for you? I don't have any official numbers, but it seems like many people can type 30+ words per minute with very little learning. Faster with practice.
Currently no chording for numbers or programmer symbols. Basic punctuation marks (,.:;?! etc.) are predicted at the end of words. Caps Lock is the current toggle switch, but I'll be adding a custom option in the next version.
As an example, tapping the keys "tges" with your left hand will output "this". See how the keys correspond spatially on the keyboard? [G -> H] [E -> I]
My favorite part of the video is where they explain all the cool things you can do ... WITH A KEYBOARD.
Seriously, though. I've experimented with chording keyboards in the past. Conceptually they're quite cool, but their advantages are outweighed by the finickiness of having to master a new set of split-second timing issues. I can only imagine having zero tactile feedback on top of all that. (And then opting for zero visual feedback as well?)
It frustrates me that there's no mention of Android. This could have been a working demo with an APK provided. I'm not sure about the overlay mode but it may be possible.
edit: As an update to this, it looks like it's possible the keyboard could be set to a 0 height View, and the Candidates view could be used, as that does not cause an application to be resized.
I don't think any method of input that relies on auto-complete, suggestions or predictions can ever reach the utility of a physical keyboard. Also, most people have invested years learning the normal qwerty keyboard, I don't see how people would be willing to adopt anything new unless it's better in all cases (including corner cases like <>!/\{}[]), or they have absolutely no choice (T9).
The only way I could see any improvement for touch screens is if you could track finger tips while they're hoovering, so that you get more information to disambiguate what the user is intending.
Autocomplete is a bugbear of mine, since it's the fastest way to type on a touch keyboard and people have a bad habit of not proof reading stuff that they are sending.
It seems that at least every few days I get an Email like this.
Well, that is certainly one of the main problems with it. For me it comes down to the edge cases. My Apple password contains letters, numbers and special characters. It takes me about a minute to type it on the iPhone. If you have to rely on auto complete for edge cases, then you can never accomplish more than the most shallow content creation, such as sexting or 5 word emails.
I'm surprised that they emphasize the fact that the keyboard can be invisible or small. Personally I think that's the most confusing feature because it leaves you tapping on the screen, or website as in the video. That aspect to me is not very intuitive as to whether or not I'm typing currently or trying to navigate the website by tapping a link, etc. Also, if it's invisible, how do you know when the keyboard is brought up on screen? Does it appear and then quickly fade away, or does it appear at all? Both cases seem unnatural at least to me, since I'm just not used to it.
Other than that, I think it looks pretty awesome. The multi-finger gesture typing looks pretty quick and intuitive.
I think that if it can be done right, that'd be the coolest feature of it.
I imagine it'd be possible to detect that you've got all of your fingers resting on the screen in the 'keyboard' configuration, and only 'activate' the keyboard when that's the case.
Like how with an iPhone it can distinguish when you press and slide your finger to scroll from when you press your finger to tap a link.
I think it'd be a cool feature because it'd give you so much more screen space for visible content (like when you're editing a document), and less of the context-switch that happens when an on-screen keyboard pops up.
In theory, there are lots of options for skins - fully visible/opaque, completely invisible (with some sort of icon to let you know it's active), any range of transparent, animated effects like heat haze, water ripple, etc. Kind of fun to think about, actually.
Watching the video it seems that the keyboard is active as long as a text-entry widget is focused, so it works exactly like a visible virtual keyboard - although introducing a mode with no visual feedback may sometimes induce errors.
Why is "p" a non-chorded letter? Clearly the other letters were chosen because they're the seven most common letters in the English language. But letter number 8 is "h", followed by "r". "p" is way down at #19. Is it just to maintain some semblance of compatibility with QWERTY? This seems short-sighted.
I think its very good that they kept the resemblance to QWERTY, even if it sacrifices small amount of speed. Steep learning curve is the biggest problem with these and if keeping the QWERTY compatibility helps I'm all for it. At least for me it seemed easier to learn because of it.
But then when you think harder, a big design flaw is that you need to put ten fingers on this keyboard to locate the keys correctly. You will have hard time use only one hand, let alone two fingers.
I think the ideal phone/tablet keyboard would be a "transparent" system based on holding the back of the screen (edit: no real estate lost and great ergonomics).
Microsoft patented the idea a while back and among other sucky things about patents, they may simply not implementing it at all.
Interesting! This reminds me of the new Playstation Vita's rear touch sensor. I think it's definitely a step in the right direction of preserving screen estate for what its primary function should be.
Pro-tip: that's not a patent. It's a patent application.
But because that stupid site doesn't tell you the publication number, it's impossible to know (without doing some searching) what's happened in the meantime. Maybe it's issued as a patent, maybe it's still pending, maybe it's been abandoned. Also, no mention of the claims? Amateur hour.
Indeed, they are called Stenotypes. Court reporters write on them at rates of 180 wpm, real champions reach twice that. But this ASETNIOP that's sold here says that experienced users reach just 80 wpm. What's up there, is it poor design?
Not... quite. Court stenographers only record speech, and it turns out that the fastest way to do this is by recording spoken syllables instead of letters and using a huge pile of shorthand to boot. Some more info here: http://plover.stenoknight.com/2010/06/steno-101-lesson-zero....
Also, the chording behavior has a timeout between keypresses. So if you type "a" & "s" quickly (intending them to be separate), it is interpreted as a single "as" chord. You have to wait a bit after each keyup.
I find that using that two finger combination for backspace is tedious, particularly because there are no arrow keys to correct errors without repeatedly pressing backspace.
Speaking of "discouraging use of the arrow keys", Imagine using ASETNIOP with vim: Despite the combo probably being a disaster in many ways, vim's use of y and p might ease some of the copy/paste issues.
At least we(QWERTY, Dvorak, Colemak, and ASETNIOP EDIT: Not the french though...) can all agree on how to press the letter 'a' (and nearly 'm').
I would point anyone looking for the perfect non-chording keyboard to the kinesis advantage, regardless of your favorite layout. http://www.kinesis-ergo.com/advantage.htm
Very cool, but the fact that the keyboard on the screen doesn't match the output (input?) nor the asetniop 'idea' makes it incredibly confusing and was enough for me to say 'ok, I'm done'. Unless it's just a problem that I'm experiencing, it doesn't seem like the author (or anyone) ever tested this tutorial out.
The demo could be done better for sure. A few steps in I think I encountered a bug that was chording keys that I was only pressing one at a time (forcing a backspace when I wanted the literal 'st').
I was working on the versions for Colemak and Dvorak users at the same time and forgot to update the original; it's sorted now. Please give it another try!
Depends on whether your keyboard is mapped to the Dvorak layout or Qwerty (basically, try one, if it doesn't work the way it's supposed to, use the other one)
Press enter about 15 times to skip the first part of the tutorial. The labels in the later part are correct and also show the other secondary chorded keys.
Most people have a hard time with typing stuff into an iPad as it is. While you can probably type easily with a concept like this, getting users to understand the chorded keyboard will be a major challenge. There are folks that don't even get swype and the dictionary keyboard. If you can get over that bump, the product has potential.
I would not hide the keyboard for any reason. It can confuse people pretty easily. If I try to click a button on a website and end up typing a character instead, it can drive me nuts.
It's...the home row for Colemak: http://colemak.com/ At first I thought they somehow figured out how I remapped my keyboard, but it's still treated as qwerty.
Sorry about that - I was working on the CHORDMAK (Colemak-based) version at the same time and forgot to fix it. It's sorted now - thanks for the heads up!
For what it's worth, the first mouse was intended to be used in conjunction with a chorded keyboard when Doug Engelbart invented it in 1963: http://dougengelbart.org/images/pix/img0030.jpg