3 ms·

I wish all emojis just disappeared. They're stupid, don't add any value, waste unicode space and developers' time.

Freak_NL 10y ago
> waste unicode space The desirability of emoji aside, Unicode has plenty of space for this. The bulk of Unicode code points are CJKV (Chinese, Japanese, Korean, and Vietnamese ideographs) characters. Compared to that, a few emoji outside of the BMP don't really amount to much. In fact, I am under the impression that the popularity of emoji had an unintended benefit. It seems to have contributed to a much better support of non-BMP¹ Unicode characters across the board. Until the extended emoji set came along, most characters outside of the BMP where fairly specialistic or extremely regional, so bugs didn't show up as fast as they do now. It used to be a tiny miracle of you managed to view a PDF containing non-BMP characters properly, nowadays that is just how things work. 1: https://en.wikipedia.org/wiki/Plane_(Unicode)#Basic_Multilingual_Plane https://en.wikipedia.org/wiki/Plane_(Unicode)#Basic_Multilin...
masklinn 10y ago
> They're stupid That's literally just your opinion, users find them useful and convenient. > don't add any value 1. they provide user value, that's why they were integrated in smartphones in the first place 2. they provided significant technical value in forcing developers to correctly handle astral codepoints, which they previously had no incentive to. Case in point, MySQL only added support for astral characters (at all, not even correct handling just storing them) in 5.5.3 in 2010, the same year emoji were added to Unicode and about 2 years after emoji were added to iOS The new crop of "variable" emoji is doing the same thing with composite grapheme clusters (aka that codepoints don't map 1:1 to user-visible "characters") > waste unicode space Utter bullshit. As of Unicode 9 there are 1085 "emoji" (including back-specified dingbats and the like going back to Unicode 1.0, according to current specifications Unicode had 139 emoji in 5.2, before japanese emoji were formally added in 6.0) in a codespace of 1114112 with 128237 allocations as of 9.0. By comparison, the Tangut block added in Unicode 9.0 contains 6125 codepoints, and the separate Tangut Components block contains an additional 755. There are no extant Tangut users and we have no knowledge of the script having been used after 1502. > developers' time. Tough shit?

> waste unicode space The desirability of emoji aside, Unicode has plenty of space for this. The bulk of Unicode code points are CJKV (Chinese, Japanese, Korean, and Vietnamese ideographs) characters. Compared to that, a few emoji outside of the BMP don't really amount to much. In fact, I am under the impression that the popularity of emoji had an unintended benefit. It seems to have contributed to a much better support of non-BMP¹ Unicode characters across the board. Until the extended emoji set came along, most characters outside of the BMP where fairly specialistic or extremely regional, so bugs didn't show up as fast as they do now. It used to be a tiny miracle of you managed to view a PDF containing non-BMP characters properly, nowadays that is just how things work. 1: https://en.wikipedia.org/wiki/Plane_(Unicode)#Basic_Multilingual_Plane https://en.wikipedia.org/wiki/Plane_(Unicode)#Basic_Multilin...

masklinn 10y ago

> They're stupid That's literally just your opinion, users find them useful and convenient. > don't add any value 1. they provide user value, that's why they were integrated in smartphones in the first place 2. they provided significant technical value in forcing developers to correctly handle astral codepoints, which they previously had no incentive to. Case in point, MySQL only added support for astral characters (at all, not even correct handling just storing them) in 5.5.3 in 2010, the same year emoji were added to Unicode and about 2 years after emoji were added to iOS The new crop of "variable" emoji is doing the same thing with composite grapheme clusters (aka that codepoints don't map 1:1 to user-visible "characters") > waste unicode space Utter bullshit. As of Unicode 9 there are 1085 "emoji" (including back-specified dingbats and the like going back to Unicode 1.0, according to current specifications Unicode had 139 emoji in 5.2, before japanese emoji were formally added in 6.0) in a codespace of 1114112 with 128237 allocations as of 9.0. By comparison, the Tangut block added in Unicode 9.0 contains 6125 codepoints, and the separate Tangut Components block contains an additional 755. There are no extant Tangut users and we have no knowledge of the script having been used after 1502. > developers' time. Tough shit?