Hacker News Reader

Comment by riskable | Hacker News Reader

It was neat to see Thomas Ran name-dropped in a New Yorker article. He's one of the most brutal keyboard reviewers on YouTube (and quite hilarious). I sent him some of my 3D printed key switches not too long ago and he did a teardown video:

https://youtu.be/H_Ym9528awM

Many have found the mechanical keyboard hobby to be quite deep... When you reach the point when you're ready to start designing your own key switches and actuation mechanisms contact me and I'll give you some advice :)

If you're curious what a 3D printed (analog hall effect) keyboard using 3D printed switches (and keycaps) looks like:

https://gfycat.com/firsthanduniformeasteuropeanshepherd

...and it has some interesting features:

https://gfycat.com/costlyglaringhyracotherium

View on HN

His reviews are great, but I find it mind-boggling that Thomas Ran (Chyrosran22) doesn't touch-type when he demos typing. e.g., https://www.youtube.com/watch?v=B9LcLgcCmus&t=644s.

Is he able to touch-type but simply chooses not to? Or did his obsession with keyboards never lead to him learning how? It's a mystery to me.

He’s said that he thinks his method of typing works fine and doesn’t lead to repetitive stress injuries, so he prefers it. I don’t remember which specific videos he said that in off the top of my head, though.

Um, excuse me what? That is touch typing. What you're thinking of is "home row", a specific method of touch typing. I also don't use home row, but I'm faster than a lot of people who do, and it absolutely counts as touch typing because I don't have to look at the keyboard at all.

Were you able to tell by watching the video if he was looking at the keyboard or not? If yes, how? Because the video only shows fingers hitting keys. There’s nothing about what is being typed or if the typist is looking at the keyboard.

>If yes, how?

I have not looked at video at all, but as a touchtypist I have a struggle not to set my index fingers on f/j.

I have seen some Internet touchtypists who claim they can use something else as keys to figuring up where their fingers are located (for example, lShift/rShift or Lenovo's clit) but in real world I have not seen any touchtypist who is not sticking his index finger on f/j constantly.

If you can type at any speed while pecking at keys you must have muscles made of pure lightning.

That said, what he does certainly looks like either an imitation of touch typing while pressing random buttons, or a method of touch typing that I’m not aware of. It certainly looks too fast for him to be checking every letter he types.

It's not hard to learn to touch type from having some reference even without al\ fngers or both hands available.

My error rate is pretty high due to recently switching keyboard software, but I can manage about 40-50 wpm on a phone keyboard and my main reference points are the corners.

Interested in what you mean by your first sentence. What's a high speed in your mind? I can type prose at about 90 WPM on a Macbook with basically just my index fingers, without looking.

What I see as pecking is having two index fingers tap down on the keyboard after locating the required letter.

I’d be amazed if you can do 90 wpm with just your index fingers regardless of strategy though. With two fingers and an average of 4 letters per word you need to tap 90 / 60 = 1.5 * 4 = 6 / 2 = 3 letters per second per finger (more if your words are longer).

I trust you can do it, but I can’t imagine what that looks like.

In my book, touch typing is being able to type without looking at the keys.

Technically it is possible to touch type with a single finger.

Of course, the more fingers you use, the faster you can type, probably needed to reach > 100 WPM, but I tend to spend much more time reading and thinking than typing usually, so in my case this is marginal.

I used to work with a guy with an impressive collection of highly customized mechanical keyboards who couldn’t touch type. I found it baffling, but the reality is that obsessions find you, not the other way around.

I think this is it. Some people are into owning fancy sports cars that they don't actually drive, building high-end gaming computers but not playing games, or buying vintage cameras but not using them to take pictures (note: I'm guilty of #2 and #3). It's basically being a collector who appreciates what's special about something, even if you don't actually use it much.

Huh, I thought touch-typing refers to not needing to look at the keys. How can you tell he isn't touch typing by just looking at the video?

> When you reach the point when you're ready to start designing your own key switches and actuation mechanisms contact me and I'll give you some advice

Any ideas on how to approach half-sized key switches? So far I'm erring to what @pseudoku3 has done by cutting down existing key switches: https://www.reddit.com/r/ErgoMobileComputers/comments/v0g89d...

By changing a few settings my 3D printed Void Switches can be cut down to 18mm from the bottom to the top of the keycap. I just uploaded some pictures so you can see how low profile they can get:

https://imgur.com/a/GejwL34

You could probably shave off another 1-1.5mm by printing some parts in resin (or switching to a teeny tiny nozzle) but I don't think that's enough to warrant the complexity or difficulties it would introduce during assembly.

For going shorter than that you can cut down the travel (the switch in the picture is 2mm of travel if memory serves), use a shorter keycap/stem, or go with a completely different mechanism (I have a few ideas but haven't had a chance to try them out yet).

What printers do you use to print the keycaps? I tried this before but it takes a lot of work to sand the surface in order to remove the layer lines.

I would imagine resin printers would be far superior for key caps compared to FDM.

A while ago I tried using CurviSlicer to FDM-3D print a keycap. It modifies G-code to make slightly-curved top surfaces.

The top of the keycap was nice, and wouldn't have needed any finishing. However, the slicer messed up the sides/insides of the keycap. Someone more motivated could try applying the curviness only to the 'dish' part of the keycap.

Curvislicer: https://github.com/dhardy35/curvislicer

I use a Prusa MK2.5s with a 0.4mm nozzle most of the time. For fine detailed legends I use a 0.25mm nozzle. Here's some keycaps I printed recently:

https://imgur.com/a/Ge0tlc5

The key (haha) to printing great keycaps is to print them on their sides. That way their top surface is smooth (to your fingers) right off the printer. The keycaps in those images have not been post-processed in any way. That's exactly how they looked when I took them off the printer.

Very cool! I’ve been idly wondering how I could build a keyboard with smaller keys, for someone I know who is a little person. The 3D printed switches seem like a great option. It looks like some of the keys in that review were smaller than standard, how far could they be scaled down? Do you have a link to print instructions you can share?

fyi, had a post on modifying existing keys to be smaller here: https://www.reddit.com/r/ErgoMobileComputers/comments/v0g89d...

The default switch width/length is 18mm to fit inside standard Cherry MX key switch cutouts (in a top plate) but you can shrink that down to probably 14mm (just change the parameters in the .scad). Alternatively, you could just cut a sheath-shaped hole in the top plate where you want the switches to go and then you wouldn't even need a switch body and could cram the switches really close together.

I used a trick like that with this silly fidget toy I made a while back:

https://gfycat.com/calculatingthankfulgoldeneye

Cool design!

Agreed. Very cool use of magnets

Reply on news.ycombinator.com