Jim ThompsonIf you look at Morse code, alphabetic characters are represented by one to four units, each of which can be either a "dit" or a "dah". In traditional telegraphy with a single key, "dits" are made by short taps, and "dahs" by longer ones. But devices like "iambic" keys, use two buttons, one for "dits", and the other for "dahs".
The problem with Morse code, though, is that it is serial. You have to key the units one after the other. But by using multiple fingers, whole characters can be keyed in a single stroke. On the left hand, the first unit is keyed by the pinkie finger, the second by the ring finger, then big finger, then index finger. The units are right-justified, so that the shorter characters use only the strongest fingers. On a right-handed version, the units would be left-justified.
The really GOOD thing about Morse code, is that the most-used characters use the least number of units, making them easier to key. And among characters using the same number of units, the most common ones are either all "dahs" or all "dits", which again are easy to key.
My first question about this is, do I really need a wearable device? Because another option is to just make a two-row keyboard with a "dits" row and a "dahs" row.
Yes, and no. Yes, because wearable (along with Bluetooth) means being able to key in text on a cell phone, instead of using a ridiculous touch screen keyboard. And no, because I don't spend most of my time on a cell phone. I spend most of my time on a laptop. And my primary issue with laptops (or desktop computers, for that matter) is having to shift my right hand to go from "typing mode" to "mousing mode", especially when doing any kind of graphical work. So really, all I need is a keyboard I can use with one hand, leaving the right hand available full-time for the mouse. It doesn't really even HAVE to be a chording keyboard. I currently use hot-keys as much as possible just to avoid having to move my hand to the mouse, and pretty much any left-hand-only keyboard would fix this. I've even written applications that have all of their hot keys reachable from the left hand, specifically for this reason. So if it turns out I can't come up with a wearable solution, that wouldn't be the worst outcome.
So here are the objectives, in order of importance:
1) Use the left hand only.
2) If chorded, use a single stroke to generate most characters.
3) If chorded, use Morse code, if possible.
4) Be usable while sitting, standing, or laying down.
5) Connect to both cell phones and computers, using Bluetooth.
Next question: Morse code? Really?
Well, yeah. My first notion was to use Baudot code, which happens to still be in use by some amateur radio operators. I assumed that this code would (like Morse) use the simplest combinations (i.e., single keys) for the most common letters, and progressively harder combinations for less-used letters. Which it kind of does, but Baudot is a code completely unrelated to anything else, and would certainly involve a lengthy learning curve. But the primary reason it's not suitable for a chording keyboard is that the figures codes in the standard ITA2 flavor of Baudot were assigned based on the arrangement of keys on a typewriter, so the codes for 1, 2, 3, 4, and 5, are the same as the codes for Q, W, E, R, and T, so the codes for numbers aren't even in sequence. Morse code, on the other hand, really does make the most-used characters ultra-simple to key. Special characters require more than four units, which is a problem. In fact, numbers are a problem, since they are all five-unit characters. So I may have to relax the objective of making all characters with single strokes. There are two ways to do characters longer than four units that spring to mind:
1) Use a Baudot-like method of having shift codes to go from letters to figures. This is good for numeric entry, since it stays in numeric mode until you switch back to letters. And there can be multiple shift sets, so you have one for capital letters, one for numbers and common punctuation, and another for less-used symbols. Kind of like how the small keyboard community does it.
2) Use a thumb button to tie multiple strokes together. For example, a Morse comma is --..--, which is six units, but it can be thought of as Z and M mashed together, or G and W if you prefer. So the thumb would be held down while "GW" is keyed with the fingers. This also solves another problem, which is, how do you type a space? That would just be a single tap of the thumb, with no fingers pressed. I think that numbers would be fairly quick and easy as well. A 1, which is .---- , would be keyed as . followed by ----, which is pretty easy. 2, which is ..---, would be keyed as .. followed by ---. Also easy. And so on. Some codes would have to be invented, for the characters that aren't included in Morse, but these could be made of letter combinations that remind you of the character, such as "LB" (.-..-...) for "#", and unlike real Morse code, these combinations wouldn't have to avoid conflicts with the existing letters. For example, if you want to code ">" as "GT", if this were real Morse code, that would be --.-, which happens to be the letter Q. So there would be characters here that couldn't transfer to serial Morse code transmission, which is something to keep in mind.
The good news here, is that this isn't an either/or thing - both methods could be used to extend the code map.
Furthermore, I still believe that Morse code has a place in the world, and that place is in ultra-low power, ultra-low bandwidth communications. So I still have some motivation to learn it. And I think this would become easy if I was using it for everyday tasks. I don't learn very well when I'm just doing exercises - I spent a whole quarter in college solving differential equations, but I couldn't solve one today without some serious review, because none of those equations were useful to me. They didn't help me accomplish anything. Same thing with German. I studied German for four years in high school, but at the end of that, I couldn't have held up my end of a conversation with a five year old. I DO learn well when I'm accomplishing things with what I'm learning. So if I can answer emails in Morse, that might be what finally works.
Next question: how hard is it to mount switches on fingers in such a way that two different tapping angles can be detected reliably?
I've already tried keying the Morse character codes on a conventional keyboard in the manner described here, so I know that it's easy enough to hit four-finger codes on two rows of keys that are spaced 19 mm apart, both laterally and .. whatever that other direction would be called. Not "longerally", certainly. Synonyms.com suggests "mesial" as an antonym for "lateral", but "mesially" doesn't seem to mean what I would expect, nor would anybody else know what the hell I was talking about. But as I was saying, I can test the ease of learning this just by writing a program that translates multiple keypresses into the appropriate characters, using 1, 2, 3, and 4 as the "dah" keys, Q, W, E, and R as the "dit" keys, and V, B, and maybe N for mode shifts with the thumb, (I can't use QWER and ASDF, because these leave the thumb with nowhere to go but the spacebar. This is unfortunate because the lateral shift between 1234 and QWER keys is twice as much as that between QWER and ASDF.) So back to the actual question, if two switches can be mounted such that one is activated when the finger is pressed directly down, and a different one is activated when the finger is extended by 19 mm and pressed, then making this a wearable thing is probably doable.
Next question: is there already something like this out there?
Yes. Several, but I don't see any using Morse code. Although this could just be a matter of programming. The Tap Strap, for example, allows you to completely customize your tap patterns. The most significant thing I found, though, is that a two-keys-per-finger wearable device has not only already been invented, but was patented by John W. McKown, in 2002:
https://patentimages.storage.googleapis.com/b4/6a/9f/5d67978400acbb/US6429854.pdf
See his work at http://www.chordite.com/.
His device is wearable, and as I said, accepts two different finger motions to get three states out of each finger (up, down flat, and down curled), which is quite ingenious, in that it even allows you to type in thin air, without having any surface at all to type on. And his approach might actually work for my idea, which is unfortunate, since I wouldn't be able to sell it, or anything similar to it. Maybe. I mean, that WAS 17 years ago, so the patent will expire relatively soon. Something I did not know about patents, is that in the U.S., you must pay big money (to my way of thinking) at the 3.5, 7.5, and 11.5 year marks after your patent is granted, in order for it to remain in effect for the full 20 years. And according to the U.S. Patent and Trademark Office, Mr. McKown made his 11.5 year renewal payment in 2014, so yeah, it's still in effect.
BUT, being able to type on thin air and type while walking around is not one of my objectives. Typing while walking doesn't seem like that smart an idea anyway. A two-row keyboard could be made that is somehow usable in two different ways, one of which is on a flat surface, and the other is against some other part of the body, such as a thigh, a belly, or the other forearm. Or even the forehead. No. Not the forehead. Something flexible, like a silicone mat with indentations for the fingers and embedded switches or piezo sensors might do the trick. Or a rigid board could stil be used, if it can be elastic-strapped to a forearm or thigh, or operated inside a pocket.
And fingertip-mounted switches may still be the solution. That would be the next thing to test once the whole Morse thing is determined to be viable. I don't mind having to strap it to my palm, with the electronics on the back of my hand. Stealth is not one of my objectives - I'm not going to be using this at the blackjack table.