Engelbart's Violin (2012) | Hacker News Reader

Engelbart's Violin (2012)

loper-os.org

28 pointstosh10 comments

"a computer system should maximally reward learning. (The way Emacs does.) And that this should certainly be true of a computer one uses for 8-14 hours a day, for decades."

I think this is a good take, modern interface design focuses on instant gratification/short term thinking and ease of use more so than on giving people good levers, or mind bikes. I think the truly brilliant tools enable use and creativity beyond their initial purpose. Sure, that stuff only becomes relevant when you're pushing the envelope and most people won't be trying to do that; but I believe focusing on that edge has long term benefits that are hard to quantify.

I'm not sure about any of the keyboard stuff, but that particular idea I resonate with.

Some past threads:

Englebart's Violin (2012) - https://news.ycombinator.com/item?id=11997761 - June 2016 (20 comments)

Engelbart's Violin (2012) - https://news.ycombinator.com/item?id=8510129 - Oct 2014 (27 comments)

Engelbart's Violin - chorded keyboards - https://news.ycombinator.com/item?id=4016421 - May 2012 (38 comments)

A lot of things about NLS were wrong; this chording keyboard was one of them. Typing speed never went above 40 wpm, because each hand stroke was a byte, a character; QWERTY can hit triple that. Alternative chording keyboard designs like the Stenotype design used by Plover can achieve even higher speeds, but I don't think the Microwriter did.

However, "The emphasis is on usability – without the necessity of training. The exact opposite of Engelbart’s approach," is correct, and it does explain why Engelbart's career went downhill in the 01970s and never really recovered.

Engelbart's chording system wasn't faster than QWERTY, but chording systems of the stenographic sort are the fastest way to get fingers accurately typing words, allowing over 200-300WPM for mortals. A major source of error with QWERTY is timing. We type "teh" instead of "the" if 'e' lands 1/100th of a second before 'h'. Or we type "THe" when we meant "The" because the Shift key is error prone to hold and release just right.

Piano is known to be very hard but it forgives keys being a few milliseconds off because the ear can't hear if a C-G chord was actually C-3-milliseconds-before-G or G-3-milliseconds-before-C. (The wavelengths of the notes themselves contribute to the latency of the ear precisely determining the order each note was struck.)

So stenographic chording systems help eliminate the error of having to exactly time the ordering of key-up and key-down events. If laptops shipped with steno keyboards they'd be faster and more pleasurable to use but "good enough is the worst enemy of best" I suppose. Shakespeare and Tolstoy didn't need high speed text input systems to write all their masterpieces. Modern computerized systems for writing musical scores haven't given modern society an abundance of Beethovens and Mozarts cranking out works like the 9th Symphony and the Magic Flute at 100 measures per minute.

Funny, as I write this comment my thinking is beginning to evolve to consider that maybe we would be better off using Engelbart's speed limited 40WPM system after all? If the effort to post a comment on political Twitter were higher maybe we'd have higher quality political discussion thereon? If I had to write this HN comment long hand with quill and ink I think I would have been compelled to get my point across with half as many words....

Stenography is weird though. They are typing sounds not words, presumably because there's no opportunity to fully hear, understand, and transcribe words... to be that fast you actually have to cut out comprehension. It's interesting but doesn't feel translatable.

The point wasn't typing speed, the NLS workstations had full keyboards for bulk typing. Where NLS combined with the chording shined was in editing and presentation (look at what they showed in the demo), and the NLS command set design facilitated that.

What I think Engelbart understandably missed is that 1) we are all beginners sometime and 2) we no longer sit in only one app all day so the time efficiency amortization benefit he expected is a bit worse. GUI standards claw back some of the efficiency loss (remember pre CUA?) but don't take you to the levels he was shooting for.

The intro scenario is actually a pretty accurate representation of what it’s like to try to be a professional musician or composer. Hardly anyone succeeds; most have to work at a nonmusical day job, and many of those who do succeed in supporting themselves as musicians end up in studio or backing roles with little creativity.

I am beginning to work on a handheld input device to solve these problems. Email me if interested. nasharp@outlook.com

This article is confusing typing with programming.

Professional, really, really high end typists exist--such as court reporters--and they do have very sophisticated, special made tools for typing unbelievably quickly.

But most programmers' quality of work is not limited by day-to-day issues with the keyboard. It is all about thinking.

There are custom keyboards out there, and some are very nice to use--I swear by my not-every-day-but-pretty-common microsoft split keyboard. But the idea that we need some amazing speed-optimized keyboard isn't accurate.

Half right. The job is indeed thinking. But to think we need to interact with the code, the spec, the documentation, and the other devs. All of these often require using tools on the computer and the interface and interactions of these tools can be crucial to effectiveness.

The chording keyboards can, with the right software, enable greater efficiency at these other tasks, which is what Engelbart was trying to show.

Reply on news.ycombinator.com