Voiceless typing on Google Glass is an interesting problem, but so far there's not one complete solution.
As Google Glass explorers continue to find or invent new uses for their wearable computer, the need to type has occasionally come up. This is an interesting problem, because everyone using Glass already has a smartphone in their pocket that is already perfectly capable of discrete typing though all of the same services Glass offers. The problem comes in when you're decided to post a photo from Glass, or on the off chance that you'd need to enter a password or manually enter in a word that Glass doesn't understand when you speak it. There's currently no official solution, but several groups are attempting to solve this problem by adding virtual keyboard to the Glass UI. Recently the guys at Minuum released a video that details how they would implement such a solution, but there's also a test going on right now through some students at Cal Poly - San Luis Obispo that paints a clearer picture of the challenges this kind of input will face as it moves forward.
The Minuum video did a great job of making the idea of typing through Glass look like something that you might actually want to do. It demonstrated a rapid typing system using head tracking and tapping alongside their already incredible software keyboard to create a complete thought. The Cal Poly test helps bring this back down to reality a bit. Their test APK can be installed on any Glass unit, and it walks you through three different keyboard types with two different assistive technologies. You can either swipe around from letter to letter and individually peck in your words, or you can use head dracking to move around and just tap the touch panel on the side when your reach the letter you want. The three keyboard layouts were T9, traditional computer keyboard, and a circular keyboard. Each of these layouts offered distinct strengths and weaknesses when using them, but they all had one thing in common. None of them were even remotely enjoyable to use.
Each keyboard made it possible to type things directly on Glass, and in most cases I gradually got faster as I used each mothod more. By the end of each typing test, however, my arm was tired and I had accomplished very little. Entire sentences took nearly a full minute to type, and passwords that required special characters (which, in my opinion, they should all have) meant flipping back and forth between keyboard lauouts to find. It's like using the keyboard on your phone right now, only ever character takes up most of the screen and you have to either swipe through the alphabet or tilt your phone to get to the next letter. It's doable, but tiring and ultimately not worth it. Minuum's angle is that their guesswork and prediction engines will probably cut out a lot of the nonsense involved here, but that does very little to address password input.
For now, I remain unconvinced that putting a software keyboard directly on Glass is the answer. It seems like allowing some kind of cross device input from your phone would be a better way to address this problem, but that creates a dependancy between the two devices. If Google's long term plan is for Glass to be a standalone device, this is obviously not a long term solution. Whatever the answer is, I don't think we've stumbled across the right answer just yet.