Drawille: Pixel graphics in terminal with Unicode braille characters | Hacker News Reader

Drawille: Pixel graphics in terminal with Unicode braille characters

github.com

150 pointsIvoah37 comments

Did anyone even stop to appreciate the irony of using a thing created for blind people to make something that will only ever be useful to people with sight, if at all?

Why couldn't the blind appreciate this software?

Assuming this is a serious question, I think you will know the answer if you look at how any braille display works. They basically "display" one line of text in braille at a time.

So unless you can appreciate your pixel graphics one printed line at a time, it's unlikely that any blind person will be able to get any benefit from this application.

Existing one-line braille displays don't understand braille unicode anyway. What's to stop a blind person from making a braille display with a greater display area?

I'm not the one who downvoted, and I don't know how many blind people make their own braille displays. It seems unlikely to me. (Not impossible. Another quick search reveals that multi-line braille exist, so it has been done.)

TBH I just did a quick search for braille displays to validate my assumption, and when the first few results came up looking roughly like the one-line display that I expected, I wrote this comment. I'm not an expert in this kind of thing though.

My understanding is that braille displays are quite mechanically complex, as each cell needs to be able to individually raise and lower six pips. An 80-character, single-line terminal display would require 480 tiny actuators.

Except they already make bigger than that. A 25x40 display.

https://www.wikiwand.com/en/Refreshable_braille_display

    A full page braille display with 1,000 cells (25 lines of 40 cells) was 
    developed in 2015 by the Tactisplay Corp.[3] With total 12,000 pixels
    configured in 120*100, it can show any BANA compatible braille 
    graphic page in 8 seconds. This video shows operation of the device.

http://www.tactisplay.com/product/tactisplay-table

Why did you link to this instead of the actual article?

https://en.wikipedia.org/wiki/Refreshable_braille_display

That site is proprietary, adds trackers, etc. -1

Because blind people can't see animations made with dots even if those dots were originally meant for the blinds.

Even if you were to print those on a braille display, it would be meaningless letters over and over.

Wouldn't this library be useful for rendering graphs or diagrams in a braille embosser / printer? https://en.wikipedia.org/wiki/Braille_embosser

Not sure if that's what Sir_Cmpwn meant, but that was my first interpretation of the question.

I reject this idea that a blind person is too stupid to recognize that the dots aren't representing letters when they use this software, and instead appricate it as a kind of art.

The braille displays you typically encounter will only display one line of braille at a time.

You are telling me that a blind person will be able to remember the entire line of gibberish as the geometric shape gets "printed" line by line and then remember those at 30 frames per seconds?

Come on, it's not about being too stupid to understand. It's about this technology having nothing at all to do with the blind except for hijacking useful Unicode characters.

I can look at ANSI art without being baffled by meaningless letters. I'm inclined to credit blind people with the same ability.

To be the one who takes things too seriously: blind doesn’t mean one can’t see; it just means their vision is bad enough to satisfy some list of conditions. People without sight just happen to fit into that category for some weird reason.

In before "cultural appropriation."

On a similar note, I wrote a library for drawing (24-bit) color raster graphics in the terminal if anyone's interested:

https://github.com/fenwick67/term-px

Demo / image utility here:

https://fenwick.pizza/term-px/

i wrote something very similar, but in c++

https://github.com/gciruelos/seurat

I can't wait until people rediscover Sixels and get actual graphics in the terminal in a network-transparent fashion. Modern terminal emulators support it, it just isn't widely known.

Really, is Sixels support that common now? Do you have a supported Linux terminal that you recommend for it?

xterm supports Sixels, and I'd be surprised if it were the only one.

And this library enables things like video and X11 using Sixel, plus it has a list of terminals which support Sixels in the README: https://github.com/saitoha/libsixel

Wow, sixels looks amazing. I'm honestly shocked

I guess Tektronix graphics isn't supported by anything anymore? http://physics.oregonstate.edu/~landaur/nacphy/coping-with-u...

What is old is new again. Good ol' xterm has had Tektronix emulation for around 30 years.

MacTerm implements TEK 4014 and 4105 graphics (MacTerm is my project). Support for TEK graphics does seem rare though; aside from "xterm" and "gnuplot", I can’t think of anything else that uses those commands.

Drawille works in MacTerm too; I ran some of the example programs just now. Neat, especially the rotating cube. :)

https://news.ycombinator.com/item?id=15549366

Quite reminiscent of the Teletext 'alpha mosaic' characters, which were 2x3 block graphics characters.

It could be used to make pages like e.g. https://al.robotfuzz.com/~al/teletext/c4/1995-08-26-0046.0/4...

Any program that can output this?

Any BBC Micro emulator would do the trick.

My previous attempt at a pithy complaint pointing out the license of this project was flagged, so I will attempt to provide a bit more value:

This is a great idea and would be nice even just as an art project but unfortunately I can't do much personally useful with it.

The choice is 100% up to the author and I respect how they avoid losing control of their IP. It is interesting to see the uptake and I wonder how many project using this tool are in violation of the license.

https://www.gnu.org/licenses/why-not-lgpl.html

This is a really neat looking project (and the GH readme references the original HN discussion). For folks who are generally into TTY UIs and dashboards, I also want to point out the Blessed terminal rendering lib and its ecosystem (which even includes a React bridge!):

  - https://github.com/chjj/blessed
  - https://github.com/yaronn/blessed-contrib
  - https://github.com/Yomguithereal/react-blessed

(edited for syntax and also to clarify I am not affiliated with the above, just a fan)

I think the block characters are easier to read, but with a 4x4 grid the resolution is lower: https://en.wikipedia.org/wiki/Block_Elements

The resulting pixel resolution is a bit screwy, especially if you use a tall, thin font like the one I tend to. Too many pixels horizontally, too few vertically. The upper block/lower block characters might be better in terms of aspect ratio, but then the overall resolution is made even worse...

The 2x3 Braille grid makes a lot more sense.

There are also 'sixels' (essentially 6x1), but terminal support is… not great. https://en.wikipedia.org/wiki/Sixel

Some cool images: https://github.com/saitoha/libsixel & https://github.com/saitoha/PySixel

This is fantastic!! This should be supported in every terminal. What a dream.

The rotating cube is especially cute. I challenge anyone to now write a clone on the classic game Elite with this.

Awesome!

I wonder if unicode characters could be used on top of libcaca for an even more accurate rendering.

I've dreamed about doing an automated "ASCII"-art converter that utilizes all of Unicode. You could accomplish some pretty crazy line art if you put your mind to it, but the analysis would be super challenging.

Also you'd either need a monospace Unicode font (which I don't imagine has anywhere near 100% glyph support) or incorporate spacing and kerning rules...

I would think about feeding rendered unicode characters to train a visual algorithm about which character may match, then "on the go" do a bitmap comparison to select the best one for the given case.

The first part could be something like shapecatcher.com where you select the top N plausible candidates, then on the final part you select the best one for the case at hand. (well at least that's how I would try to implement that first)

Anyway yes, it would be great!!

Reply on news.ycombinator.com