mastodon.me.uk is one of the many independent Mastodon servers you can use to participate in the fediverse.
Open, user-supported, corporation-free social networking for the UK.

Administered by:

Server stats:

483
active users

#unicode

8 posts7 participants0 posts today

Ça fait déjà une semaine que j'étais invité à mon premier @ParisWeb pour parler un peu #unicode merci à tous pour vos messages, discussions et intérêt ! Félicitations au festival pour ses 20 ans, sûrement la conférence la mieux organisée que j'ai pu faire. Merci à l'équipe de bénévoles aux petits soins et aux interprètes LSF qui ont fait un taf de dingue. Trop cool d'avoir rencontré @tk @joachim @tut_tuuut @yopox @marc_bouvier @sabderemane et j'en oublie sûrement d'autres désolé !

Honestly, #emoji and icons in #Unicode are a true horror.

Yeah, sure. It's great that you don't have to use <img/> anymore and you can just paste a random Unicode character. You can get graphics into fields where only text was originally intended (like bug summaries). Even better, you can now easily get cool colorful icons on terminal with almost no effort.

However, it is an #accessibility nightmare. People are now encoding *information* in random graphical symbols. Symbols that require huge fonts to render, or huge character tables to describe.

Yeah, a bare <img/> carrying information sucks. However, you can add a *meaningful* alt-text to the image, and accessibility tools can use that text to provide meaningful context. Like "bug fix".

However, emojis and icons are symbolic. The best you can get is some description like "hammer and wrench", so people can kinda figure out that it's probably a "bug fix". Or maybe it was a "maintenance task"? Or you'll get a "unknown character 0x1F6E0". And I'm sure people will surely enjoy cross-referencing a "legend" of such "unknown characters".

So... In case you where wondering, you can use emojis as identifiers in #swift, but I think most people know that....

But you can also just use runes:

func ᚠᚡᚢᚣᚤᚥᚦᚧᚨᚩᚪᚫᚬᚭᚮᚯᚰᚱᚲᚳᚴᚵᚶᚷᚸᚹᚺᚻᚼᚽᚾᚿᛀᛁᛂᛃᛄᛅᛆᛇᛉᛈᛊᛋᛌᛍᛎᛏᛐᛑᛒᛓᛔᛕᛖᛗᛘᛙᛚᛛᛜᛝᛞᛟᛠᛡᛢᛣᛤᛥᛦᛧᛨᛩᛪ() { return }

Do what you will with this information.

On Linux, how do you quickly enter #emoji and #unicode characters? I just learnt that we can do Ctrl-Shift-u and type in the Unicode code point in the browser. That's how I wrote the rupee character in the previous post. It's not gonna be possible to remember too many codepoints in this manner.

Now, I am aware of emoji pickers which easily allow you to choose emoji, but what do you all use for unicode. Some of the emoji pickers don't show the unicode characters.

Also, just realised that in WezTerm Ctrl-Shift-u it opens a helpful emoji/unicode picker. Ideally, something like this should be easily accessible from anywhere with a quick keyboard shortcut.

I'm experimenting with using #Emacs for my terminals, especially using the #vterm package. It could be a really nice environment: cross-platform, easy copy-pasting with keyboard commands only, etc.

However, when I started up the #Gemini CLI in the vterm buffer, I found that it renders quite differently. See github.com/akermu/emacs-libvte. Basically, there are soft stripes between the #Unicode block characters and the shade characters are dithered instead of being nice solid blocks.