you know I can't help but think "using custom emoji letters for an entire toot" is a worse accessibility sin than "images without descriptions"
@KS Not necessarily. I'd expect a screen reader to perform a Unicode normalisation process on the text, which would show these symbols as the normal letters.
This is only true of things like Ⓗ → H though. Not if you try to use leetspeak, of course.
@loke I'm not talking about unicode variant letters, though; I'm talking about custom emoji that just consist of a single letter (e.g.: the hackertype letters, wide letters, etc).
Those don't have unicode code points, just shortcodes (which Mastodon replaces with an image).
Without support for custom emoji, those posts just turn into a bunch of unreadable :hacker_g::hacker_a::hacker_r::hacker_b::hacker_a::hacker_g::hacker_e:, and I can't imagine it's any better for screen readers
@KS Oh wow. I see. I did not know that's what they were.
I always been against the idea of custom emoji-like symbols embedded in text. I mean, I don't like emojis much at all, but at least the are standardlised, and the symbols can be converted in an automatic way as I mentioned earlier.
These images would cause problem with any text-only medium, not just screenreaders.