Live data from Hacker News

Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

news.ycombinator.com

71–80 of 110 posts

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#71
I don't think I have the skill to learn a stenographer's keyboard, but I would love to have a small chorded keyboard just for macros.

I always liked the look of Doug Engelbart's one in the Mother of All Demos. It's very basic, but I'd be quite satisfied with something like that today. He demonstrates it at about 1:40:

https://www.youtube.com/watch?v=B6rKUf9DWRI

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#72
I have been using plover for about three years now for the majority of my time spent on the computer. I don't think I type more than 50 words a month using a regular keyboard. I still use (half) a keyboard for games, and there are some programs on Windows plover does not work with. There is an embedded steno engine (javelin-steno) so you don't have to use plover, but I have not set it up yet and just stick to using plover.

I write all my code using plover, but I am not a professional programmer. I use emacs most of the time. I use this dictionary for my symbols https://sammdot.ca/steno/emily-symbols.png, this one for typing almost any shortcut combination with my left-hand https://github.com/Abkwreu/plover-left-hand-modifiers/blob/m..., as well as this http://www.openstenoproject.org/stenodict/dictionaries/cross... for moving the cursor around and selecting text. The emily-symbol dictionary is fairly popular, and most users will have some set of dictionaries for shortcuts and movement.

Its worth noting you can type single letters, so if you don't know a word or don't care to learn it, you can still type the word out. You don't have to memorize every single word.

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#73
The best value-for-effort upgrade to keyboards for coders would be to get a keyboard where each thumb can use two-three keys each (rather than just being able to use a single giant spacebar).

Stenography looks like 'high effort, high reward'.

Whereas, bringing keys like backspace, enter, esc, tab to within easy reach of the hands on home row is going to be a big increase in comfort (and I'd be surprised if it was slower).

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#74
post #52

Earlier quoted context omitted.

Steno will get you to 200+ wpm, not 100. I've been trying to learn steno on and off for 5+ years. The problem is I already have 150 WPM on qwerty and I cannot think of a message on Teams faster than I currently type. The opportunity cost is far too large to justify unless you need to transcribe someone else's speech.

I top out around 100 wpm but still rarely bother to push myself to hit my max—and pretty much never when writing code. I can’t really relate to “I need to type faster!” programmer optimizations, nor complaints about things like static typing slowing people down because it’s a few more characters (more thinking, I’d get, but some folks do seem bothered by the extra keystrokes) since input speed is nowhere near being m…

> I can’t really relate to “I need to type faster!” programmer optimizations, nor complaints about things like static typing slowing people down because it’s a few more characters...

One common loop in programming is "hypothesize, test, evaluate".

If you're exploring or playing around, the quicker you can execute iterations of this, the more likely it is you'll succeed in what you're trying to do.

In that case, stuff like "unused import is a compiler error" or "static typing required" slows down iteration, so gets in the way of rapid prototyping (or whatever).

"Typing quicker" wouldn't benefit the 'hypothesize' nor 'evaluate' parts, sure, but it'd help you reduce the time it takes to test an idea.

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#75
post #35

In my experience, typing speed is never the issue. I've worked with truly 10x and better programmers in my life and the road block for even them is thinking not typing.

This is likely true in many instances and types of problems, where you are working on novel issue, and just need to think deeply.

On the other hand, I routinely encounter simpler problems where I can metaphorically "see" a page of code in my head that just needs to be flushed out of my mind's buffer, and I'm waiting for my hands to do the typing. I'm a moderately fast typer (somewhere between 70 and a 100 wpm), and it routinely is a blocker in my flow.

So, your mileage may vary, and I don't fully buy the argument. If one is going to be doing many hours of this activity every day for their life, why not get good at this aspect also?

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#76

The best value-for-effort upgrade to keyboards for coders would be to get a keyboard where each thumb can use two-three keys each (rather than just being able to use a single giant spacebar). Stenography looks like 'high effort, high reward'. Whereas, bringing keys like backspace, enter, esc, tab to within easy reach of the hands on home row is going to be a big increase in comfort (and I'd be surprised if it was slo…

I've used my Kinesis Advantage 2 for years and am inseparable. I'd like to swap to the Kinesis 360 but my current keyboard is just fine and it feels like a small upgrade

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#77
post #51
post #10

Earlier quoted context omitted.

Dvorak is much more comfortable than qwerty, in my opinion. I never actually cared about speed, it just feels better.

Dvorak keyboard's fatal flaw is when you have to type on someone else's keyboard. Standardization has its benefits, even if less than ideal. Trackballs have a similar issue.

To an extent, Vim and Emacs have a similar issue, especially if you spend time customising these.

Often, the benefits from using an improved tool outweigh the costs of it being non-standard.

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#78

Depends on your goal. Chording technique is superior when typing words contained in the dictionary. Meaning that typing some rarely used word required typing it multiple times to "confirm". Writing code does not suite well for this, since coding with completion contains much more punctuation than plain text. Instead, check out ergonomic mechanical keyboards: low-profile, split, with columnar stagger, preferrably with…

> preferrably with 36 or less keys.

I think it's worth emphasising that these small keyboards bring the full functionality of the keyboard to within reach of the hands resting on home row.

They aim to increase comfort by reducing hand movement, and stretching / use of the pinky fingers. This benefit comes at the cost of a (slightly) more complex keymap.

The key thing isn't "less than 60 keys", it's "two or more thumb keys on each hand".

Some of these ergonomic keyboards opt for more than 34-36 keys. (e.g. the Moonlander is relatively popular).

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#79

Earlier quoted context omitted.

I top out around 100 wpm but still rarely bother to push myself to hit my max—and pretty much never when writing code. I can’t really relate to “I need to type faster!” programmer optimizations, nor complaints about things like static typing slowing people down because it’s a few more characters (more thinking, I’d get, but some folks do seem bothered by the extra keystrokes) since input speed is nowhere near being m…

> I can’t really relate to “I need to type faster!” programmer optimizations, nor complaints about things like static typing slowing people down because it’s a few more characters... One common loop in programming is "hypothesize, test, evaluate". If you're exploring or playing around, the quicker you can execute iterations of this, the more likely it is you'll succeed in what you're trying to do. In that case, stuff…

> stuff like "unused import is a compiler error"

For things like this I’d much rather spend the time to improve “quick fix” suggestions in my development environments. The computer can fix the problem much more easily than I can learn a new keyboard layout.

Re: Ask HN: Has anyone tried adapting a court reporter keyboard for writing code?

#80
I’m a computer engineer - a lot of programming, but also a writer so a lot of writing, and we also generally do a lot of writing in our daily lives - emails, chats, prompting… :D I am generally split between English and French writing mostly.

I use a combination of Dygma Defy with its awesome thumb cluster, along with macros for frequent series of letters (think “tion” and such) as well as chords using https://github.com/rvaiya/keyd/ . And I use the Optimot layout because I’m French, for English speakers, Dvorak is probably enough, but Colemak and many other alternatives offer various advantages depending on your usage.

As many have hinted here, it’s absolutely not only about speed. It’s about comfort, both physical and psychological.

Using the thumb is a great way to avoid moving the hands too much because now your pinkies don't have to reach keys on the side which generally causes a slight extension of the hand. After a year of using the Defy, I don't have any form of strain building up in the thumbs, even though I use them quite a lot - but still a lot less than other fingers.

Macros for short series of letters are very powerful in my opinion: it doesn't necessarily goes faster as it breaks the flow of typing, but you have a lot less keys to press which also minimizes errors. The same goes with chords using keyd (I know it’s originally not exactly designed for that usage, but it still works great at least for me) - less keys, less coordination.

Finally, the choice of keyboard layout is critical. Originally I’ve switched from the French Azerty to Bépo because I notices how much my wrists were moving when I was typing in French with Azerty compared to English in Dvorak. There was a huge difference and I could feel it in my bones after long sessions of typing. So yes, choose your layout wisely.

As a last note, typing speed does matter. But it’s not about typing at the 0.001% fastest percentile. Typing speed will not make you really faster, but you just don't want it to slow you down too much. Typically, you don't want to be in the position where your thoughts go so fast that you feel the frustration of not being able to type fast enough and losing some of your thoughts in the process. Besides, for coding it’s much more about having the right tools at your disposal: powerful auto-complete and suggestions, easy refactoring processes, keyboard shortcuts to do everything you need, easy access to symbols on your keyboard/layout, etc. Typing speed is rarely an issue while programming compared to when writing plain text.

Post reply on HN