Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Yeah a lot of arguments are of the form "when Unicode can include the pile-of-poop-emoji, surely it should include <my favorite symbol>".

But the question is not if the icon is practical or silly, the question is if it makes sense as a Unicode character or should be handled at some other level like a GUI control or widget.

Should GUI icons, widgets and controls be covered by Unicode? That is a can of worms.

I understand the usefulness of the "external link" icon, but should it really be something you arbitrarily type into a text? Shouldn't it be part of the rendering of a link?



As for the pile-of-poop emoji there is a very good reason for it to be included in Unicode: it is one of the original emoji, made by Softbank in 1997.

The reason: Softbank and Docomo both used proprietary emoji sets, widely used in Japan for text messages. Not including these in Unicode would have resulted in a choice: be compatible with the rest of the world, or keep using the emojis that became part of their culture. Not satisfactory. So basically, they had to put emoji in Unicode for the Japanese to use it, and excluding Japan is not really an option if you want a universal standard.

It did open an Pandora box though.


SoftBank actually put emoji into the Private Use Area of Unicode before they were codified (just as they were in an unused area of Shift-JIS before that). The original iPhone 3GS use the PUA method.

The encoding of emoji was in order to unify Softbank/docomo/au under one set so that iPhones/Androids sold by the different carriers could send emoji among each other without relying on email translators (as had been the technical solution with feature phones until then)


It made total sense to include emojis in Unicode initially.

And a great idea to put them outside the 16-bit plane, so platforms would be forced to support characters beyond the BMP because of the public demand for emojis.

But I think going forward, emojis should be handled outside of Unicode. They are really small illustrations rather than characters or symbols.


>Should GUI icons, widgets and controls be covered by Unicode? That is a can of worms

They already got all the Wingdings symbols https://en.wikipedia.org/wiki/Wingdings


Yeah because Wingdings was a preexisting character set and Unicode was designed to be a superset of all existing character sets in use.


Thus the inherent contradiction: Unicode is for text, it must include all existing character sets which it aims to unify, and previous character sets weren’t only used for text.

The obvious hack is to make a custom character set called “Rejected by Unicode”.


The goal was to eliminate "we can't switch to Unicode, it doesn't have ® symbols" objections.


It would probably be easier just to design a png or svg icon for the UI widgets you need. Then you don't have to lobby a consortium and wait for them to accept your icon.


To offload the design to the font foundry is even easier :)


Unicode Russell's paradox


Isn’t Wingdings just a font that can work with whatever character set?


Yeah, it was basically a hack to extend the character repertoire in the time of 8-bit character sets. The kind of shenanigans Unicode was designed to replace.

You still see this kind of mischief when Outlook users type a smiley, and it is rendered as a "J" with the wingdings font. Readers of the mail where styling is stripped or who haven't the font installed just sees a confusing "J".


Unicode is full of whole blocks of weird sets of lines and borders. As far as I’m concerned adding the symbol just to make it easier to style links makes total sense.


Those "weird" characters has been used for ASCII and ANSI based text-mode user interfaces since the dawn of UI as a concept.


Being pedantic, those box drawing characters are neither ASCII nor ANSI. They are part of the original IBM PC character sets. Windows refers to them as the OEM character sets.


Unicode’s box drawing seems to go way beyond ANSI, though, and PETscii was only added very recently (IIRC).




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: