“So, why can’t we just use Unicode characters in programming language syntax?”
Graeme Coleman @graemecoleman
20 Feb 2019
Hello, my name is Graeme, I have a PhD in computing, and I am a senior accessibility consultant, but when I want to type "é" on a Windows laptop I go to Beyoncé's Wikipedia page and copy/paste the letter from there.
24
19
153
homoglyphs create mischief. If you've ever copy/pasted code and it broke b/c of a different quotation mark, you know why _code_ needs a very restricted set of symbols to work with, even as _data_ can make free use of unicode. But a text file containing both should err to code.
2
12
JS since ECMA-262 3rd Edition has allowed Unicode letters (and digits if non-initial) in identifiers. Anyone have horror stories?
6
6
22
Once D3 source code was written using mathematical notation. I was horrified when they replaced that beautiful source code with boring latin identifiers.
5
2
40
.ಠ_ಠ { --（╯°□°）╯: ︵┻━┻; } is valid CSS.
8
183
21
569
Replying to @tabatkins @bterlson @BrendanEich @foggyanabasis @TimSweeneyEpic
Using fullwidth parentheses and other things that make it look like there are spaces where there aren't is pretty confusing. Maybe it's intended to be?

Sep 28, 2022 · 12:10 AM UTC

1
Replying to @davidbaron @bterlson @BrendanEich @foggyanabasis @TimSweeneyEpic
It's certainly not meant to be *understandable*, sure. (Tho it's technically not valid, due to the degree signs which are outside the valid ident char range. But there are visual copycats I could use instead.)
1