Slashdot Mirror
Alternative Text Input Methods?
A reader asks: "I've been tentatively researching existing solutions to enter text without a full (QWERTY) keyboard, and besides touchpads, morse code, and the system used in mobile phones (added with dictionary-powered predictive methods) and some wild gesture-based ideas I've come with the following systems: chording, as exemplified in some of the keyboards in this gallery, and Thumbscript (which is patented). Does anybody know any other good methods to enter text on a limited keypad? This issue is likely to become more important as new, smaller devices of all types enter the market." Interesting question. Devices that are already in the market (and those soon to hit the market) which are designed to be portable in size (but not in functionality), this will be a huge issue. Try editing song titles on today's portable MP3 player or writing a paper on your Palm. It's doable but not very pleasant. Such alternative methods would be a welcome addition to such devices.
Why are we trying to cram effective text input onto a cell phone, PDA, MP3 player, or similiar devices? I can see the use, but I see it as more useful if we have a bluetooth or similiar system, where the cpu/main unit can be 1 inch square, and you can use mini/full size/whatever input device via short range wireless link? Added advantage is share phone books, etc, a good thing betewwn phone and pda.. that way, I can carry one keyboard for my cell phone and pda... it'd be nice to see something like that. I know, bluetooth, but I don't see it as useful yet..
bash: ispell: command not found
This sig left intentionally blank.
If you have number keys 0-9 and 7 or more other keys, then use one to toggle into hexadecimal mode, and enter the hex codes for ASCII (which you'll eventually memorize after having to look them up constantly).
I see even classic Slashdot is now pretty much unusable on dial up anymore.
A company called Infogrip makes a keyboard similiar to this. It has seven keys, four cording and three meta. It looks kinda cool, but runs about $200, so out of my price range to get one for playing with.
A 9 key method that I like, and that was on slashdot a very long time ago, is quikwriting.
The original version is similar to thumbscript(having 9 keys, each letter being made of multiple keys), but more complex(81 in thumbscript vs 128 in the simple version).
Also, since the last time I checked the site, they have also come out with a new 18 key version(two side by side 3X3 grids) that takes better advantage of a palm's writing space.
One nice feature is that it is a continous system, so if you are using a palm or something, you never have to lift your stylus.
It is free for personal use, but you have to contact them about any commercial use.
What happened was the other UNIX adage, "Everything should use similar input". That's why shells now accept "vi" editing commands, and why csh and tcsh are often frowned upon (quite powerful, but breaks the Bourne Shell standard).
Personally, I think that we will end up with two dozen different handheld input methods until one gains clearcut approval (or Microsoft makes one proprietary, whichever comes first). This will be nasty.
--The basis of all love is respect
Such an array on 6 buttons could be perfect for one-hand use and small devices, just the size to conveniently fit in a hand. Text entry speed can be quite high, provided that chords are designed carefully and logically. This does not take unusual hardware, too. Of course, learning curve for such a thing is much steeper that for conventional keyboard, because there is no visual clue what chord would produce which character.
I heard of such a thing being successfully used in military aircraft -- arguably, a fighter jet pilot cannot sacrifice both hands for rapid data entry and would hate to move hands too much at several g's. If it may prove useful in consumer electronics, is more doubtful :-)
Computers make very fast, very accurate mistakes
Pointandgrunt.net had a piece on this topic in late May called '10 big finger', all about "How do we get text into small gizmos quickly and accurately?". It is fairly detailed, and links to many different sources of information about just this kind of thing. Exactly what you're looking for!
The story can be found here.
And register for an account and get a really low user id number. I know I did. You never know, in 5 years, you might get to gloat about it... ;)
Hope this is of some help.
If anyone doesn't know what predictive text input is, basically the phone has a mini dictionary built in, and you just press the number the letter you want is on [b]once[/b], and it guesses what the word you want is. Eg: 843 would guess 'the'.
It works really well, and on devices like mobile phones, where text input isn't the whole purpose on the device, it would do just fine.
What if all of these new-fangled gadets didn't have any method of text entry on them what-so-ever? I mean how often do you have to change the song title on a MP3 player and why would you ever want to write a paper on your palm?
What ever happend to the old UNIX adage that every program should do one thing and it should interact with everything else? The same applies to these new devices. Your cell phone should do nothing more than send/recieve phone calls, your PDA should do nothing more than keep appointments and an address book and your MP3 player should do nothing more than play MP3s. Now where this comes together is bluetooth. Now you want to call someone who's phone number is entered into your PDA, no problem, just bring up the entry, hit dial and your PDA will tell your cell phone to dial that number.
No back to the text entry. What if you had one keyboard that would allow you to enter into every device out there? This could even be a portable computer. Size here wouldn't be an issue, if you want a full sized keyboard to change your MP3 song title no problem, if you want to use a small keypad with the same MP3 player again, no problem! The key here is to get every device to speak the same langauge. It all goes back to creating a standard, and sticking to it!
I think that they are trying to put too much duplicate functionality into all these devices. The focus should change to making each device do one thing, and do it well.