This was available in some (probably limited and gimmicky) form decades ago in devices like the Reflection Technology Private Eye. This is the best link I could find:

http://www.loper-os.org/vintage/paralleleye/eye.html

... I think it was in Hackers, so you know it's good.

It sounds like a good "eye-dea", but its not. Reading any real amount of text on Glass gets tiring after a few minutes. It's really best for momentary use. They really did screw up the software. Why didn't it integrate more with the phone in the first place? So much wasted opportunity, even with the physical limitations like resolution, eye strain, and battery life. I've relegated it to large events and vacations, it works great for documenting that stuff!

regarding eye-strain: do you think it's due to the specific display hardware Glass has at the moment, or a fundamental problem with headsup displays?

The position is a little tiring, but its semi-transparent, and the text is small, and the resolution is low. Also, you are reading with one eye. That makes a difference after a bit.

Thanks! Semi-transparent and one-eye would be tiring. Fixing them (two-eyes + removable patches) would be a different product, almost an occulus. But it's helpful to see the keyboard...

I go bicycle touring, and a heads-up map while cycling would be convenient. Plus, it would require less room to use in a small tent/bivvy bag.

Here is where I've found it good use: biking, hiking, parties/wine tastings/weddings, parades, etc. Driving is a fairly good use case as well. If there is a clear use for it people don't seem to care I'm wearing it, and its pretty fun with Auto-Awesome.

that's (sadly) what I keep envisioning an occulus for.

not for convenience, simply to shut out the rest of the world. full screen mode on steroids.

The mention of steroids seems spot on. I'd imagine using the Oculus for long time spans and/or demanding tasks would have so many side effects, like severe eye-strain, lack of resolution of the environment, lack of reactivity to external inputs, and so much more.

In the future, we all (developers / engineers) go to work by getting out of bed and strapping on a helmet to join our team in a virtual office environment.

earphones for your eyes

You wouldn't need an actual keyboard, of course, because the glass could just watch where you're typing and convert that into keystrokes.

Um, no. This isn't easy to do, except in movies. Glass doesn't have the CPU power to do it, so you'd have to rely on cloud power; do you really think everybody wants all their key strokes sent to Google? And language models aren't that much help with terminal commands. And even humans can't do this; you're expecting Glass to be smarter than a human... not impossible, but think about what processing power and memory it has on board, both constrained by battery. And your head would have to be in an incredibly awkward position. So many reasons, no.

Yes, I'd send my keystrokes to Google if I could do this. Sure, this won't work for everyone but it will for most. Let's optimize the problem for the 80% and continue working on the other solution.

If all the problems could be solved, it would be awesome. But I'm very skeptical. And I misspoke about keystrokes, of course: what we're really talking about sending to the cloud is finger and hand movements; the keystrokes would be coming back from the cloud. Sounds laggy maybe, on top of all the other problems I mentioned.

probably would be more effective to have an electrode connected to your brain

Why do I need to have to have a keyboard? I'd rather use voice recognition and gestures. Why do people always just want a faster horse?