And then one day there will be a generation of programmers who have only known coding via a voice-driven developer assistant. Imagine trying to interview such a person.

"Please stop shouting at the whiteboard. It doesn't work that way. I need you to pick up one of the dry-erase markers and write your algorithm to invert a binary tree."

"What? Who codes with their hands? I need my assistant!"

Would that be a bad thing though ? Assuming assistants are ubiquitous. We don't use punch cards in interviews today. Even today whiteboard interviews are sometimes not the best method to fill certain roles.

"You have to use your hands? That's like a baby's toy!"

IMHO I don't think it would fly. The most difficult things of programming are too abstract for that.

You may be interested in the guy that did all his coding with voice for a while because he couldn't use his hands, and still uses it to this day about half of the time. https://www.youtube.com/watch?v=8SkdfdXWYaI

Yes, I already know many people don't think it will work. I have that url, along with a few others in my Programming by Voice notes:

https://github.com/melling/ErgonomicNotes/blob/master/progra...

No one said that you have to do everything. Start with what's possible then iterate.

If I'm reading or thinking and leaning back in my chair I often want to say or gesture a command to scroll up or down for example, it happens more often than not when reading long articles.

(btw, if anyone knows anything about software/hardware to do that? let me know)

It would good to have the options to voice or gesture command a desktop computer.

I used to do this with Dragon Dictate. It worked very well; I recommend it.

I created speech bindings for my most common navigation tasks:

- "Down" -> presses PgDn

- "Down a bit" -> presses "down" a few times

- "Switch" -> presses Cmd+Tab (next application)

- "Backtick" -> presses Cmd+` (next window)

- "Apple L" -> presses Cmd+L (focus address bar)

Latency was good enough. Bindings have a higher priority than dictation, so if it hears one of your custom bindings: it will immediately action it instead of waiting to hear a full sentence.

Mac version was mostly only reliable for triggering keyboard shortcuts. Dictation of sentences worked pretty well, but there were (many) significant bugs that meant it was not generally viable.

I eventually stopped because of one of the following (can't recall):

- OS updates broke it - lost the license in a computer switch - hands got better

For basic gestures, an apple magic mouse is pretty good (also works in windows).

Almost the identical use case - I sometimes sit back and use gestures to scroll, swipe browser back, swipe between desktops.

An external trackpad would also do the trick, but the mouse is ergonomically nice for when you're stretching or leaning back.

Voice input is an option when other types of input are not.

In the same vein, I am learning morse code. Any resources helping with this would be appreciated.

I would like to use morse code for silent, no-look I/O on my phone. I'm not sure vibrations can be quiet enough; might have to use something special-purpose plugged into the headphone jack.

You could design a keyboard for it on Android at least.

Make the whole keyboard area a blank "tap zone" where you can just use any digit to tap text into an input field.

I looked and this has already been done a few times!

https://github.com/gitonwithit/MorseKeyboard

https://github.com/whymarrh/morse-code-keyboard

https://play.google.com/store/apps/details?id=com.Input.Mors...

For learning: Morse-it app. Not sure if it's on android, but it's a great learning tool with lots of learning methods.

This guy took the first step. Its a great watch even though its kind of long

https://www.youtube.com/watch?v=8SkdfdXWYaI

I'm sure someone on HN is seeing this right now and has already started working on it :). Especially with the advent of Google Assistant SDK and the likes, it's clearly very much possible now.

The crazy part to me is working in an office where 50 people are simultaneously saying code out loud.

Or my wife ready to kill me as I'm saying code while she's watching tv. :)

Rather than my usual tirade against open offices, I would point out that options for subvocal voice pickups exist like throat mikes. You could speak softly enough your neighbor couldn't hear you, and they also attenuate external sounds quite well.

So basically you'll be trying to think above a constant, low-level din of unintelligible murmuring. Not unlike having a brown noise generator in your office. That would drive me insane.

Well, when I say "your neighbor can not hear you", I mean that literally. Not mumbling, just talking at a sub-audible volume.

And, IMO, even mumbling would be better than the current open office constant flow of loud conversations (one of which I'm trying to drown out with headphones this very moment).

Drop the prefix keyword and then we might start talking about "smart".

Oh, and it must work locally instead of sending everything to the cloud.

Yep, there are demands you can certainly make after the problem is solved.

For now, it would be helpful to many if there were no such limitations. In other words, comments like "Make it even smarter and it can't go to the cloud then I might use it" can safely be ignored for now. There's a market for an MVP that does a lot less.