3 ms·

I'm more worried about the inflation of emoji than a couple dozen unused ghost JIS characters.

by einpoklum 4y ago

wongarsu 4y ago
Godwin's second law: any sufficiently long discussion about Unicode includes a discussion about emoji :)
- raphlinus 4y ago
  Yeah, it does seem to come up a lot more often than discussions about U+5350.
edent 4y ago
Why? Unicode isn't running out of space any time soon.
- Mountain_Skies 4y ago
  640K should be enough for anybody.
  - rossy 4y ago
    With Unicode, concerns of running out of space are absurd. There are almost a million unallocated codepoints. At the rate codepoints are being allocated, we won't run out for at least 250 years, and emoji are less than 3% of those allocations. Also, the biggest limitation on the number of codepoints is UTF-16, which we have to pray is dead by the late 2200s (when it dies, we'll have over 2 billion unallocated codepoints.)
- kevin_thibedeau 4y ago
  The encoding has gotten out of hand with compound emoji. Splitting them on glyph boundaries is non-trivial.
  - estebank 4y ago
    The functionality was always there (think ê vs ê), so properly handling glyph segmenting a string requires returning a `Vec<Vec<char>>`, which makes these emoji actually very useful: it makes it more likely that implementations will do the right thing by giving people from "predominantly ASCII" locales a "tool" to exercise those codepaths. Widespread emoji adoption is likely the best thing that could have happened to proper text handling for users outside of the anglosphere.
    - kevin_thibedeau 4y ago
      Combining characters are trivial to deal with. Some of the new emoji compounding uses things like the color squares to alter an emoji and you are forced to have a table of which emoji are dual function to know if they merge with their neighbors or stand by themselves. Nothing else in Unicode acts like that. You can't properly parse complex emoji glyphs from a random starting point because you need previous context to know how to interpret following codepoints. With combining characters you just skip ahead to the next non-combiner.
- einpoklum 4y ago
  It's not about the space, it's about the use of what is supposed to represent written glyphs for graphical effects which are becoming increasingly complex. There is nothing in principle that would limit infinite expansion of the space of emojis, and their capabilities.
npteljes 4y ago
What worries you about it?
jimmygrapes 4y ago
If Slack/Discord/etc. custom emojis get used enough, do they get incorporated into Unicode? I've seen something like 40 variants of laughing emoji, and closer to 400 variants of Pepe the Frog, and I'm not even in any "alt right" or 4chan-adjacent chat rooms/guilds where I imagine there are even more. Not to mention the countless custom anime face ones.

I'm more worried about the inflation of emoji than a couple dozen unused ghost JIS characters.

wongarsu 4y ago

Godwin's second law: any sufficiently long discussion about Unicode includes a discussion about emoji :)

raphlinus 4y ago
Yeah, it does seem to come up a lot more often than discussions about U+5350.

edent 4y ago

Why? Unicode isn't running out of space any time soon.

Mountain_Skies 4y ago
640K should be enough for anybody.
- rossy 4y ago
  With Unicode, concerns of running out of space are absurd. There are almost a million unallocated codepoints. At the rate codepoints are being allocated, we won't run out for at least 250 years, and emoji are less than 3% of those allocations. Also, the biggest limitation on the number of codepoints is UTF-16, which we have to pray is dead by the late 2200s (when it dies, we'll have over 2 billion unallocated codepoints.)
kevin_thibedeau 4y ago
The encoding has gotten out of hand with compound emoji. Splitting them on glyph boundaries is non-trivial.
- estebank 4y ago
  The functionality was always there (think ê vs ê), so properly handling glyph segmenting a string requires returning a `Vec<Vec<char>>`, which makes these emoji actually very useful: it makes it more likely that implementations will do the right thing by giving people from "predominantly ASCII" locales a "tool" to exercise those codepaths. Widespread emoji adoption is likely the best thing that could have happened to proper text handling for users outside of the anglosphere.
  - kevin_thibedeau 4y ago
    Combining characters are trivial to deal with. Some of the new emoji compounding uses things like the color squares to alter an emoji and you are forced to have a table of which emoji are dual function to know if they merge with their neighbors or stand by themselves. Nothing else in Unicode acts like that. You can't properly parse complex emoji glyphs from a random starting point because you need previous context to know how to interpret following codepoints. With combining characters you just skip ahead to the next non-combiner.
einpoklum 4y ago
It's not about the space, it's about the use of what is supposed to represent written glyphs for graphical effects which are becoming increasingly complex. There is nothing in principle that would limit infinite expansion of the space of emojis, and their capabilities.

npteljes 4y ago

What worries you about it?

jimmygrapes 4y ago

If Slack/Discord/etc. custom emojis get used enough, do they get incorporated into Unicode? I've seen something like 40 variants of laughing emoji, and closer to 400 variants of Pepe the Frog, and I'm not even in any "alt right" or 4chan-adjacent chat rooms/guilds where I imagine there are even more. Not to mention the countless custom anime face ones.