Hacker News Reader

Comment by Stamp01 | Hacker News Reader

I don't understand why this character needs to exist given that, at least according to the author, it has only been seen once in the wild, and it's semantically identical to another more widely used character.

I'm glad I'm not responsible for unicode. Clearly I have the wrong mindset for it.

View on HN

Perhaps it's relevant to look at how it was introduced - as a "package deal" with many, many characters from medieval cyrillic literature, as described in this proposal https://www.unicode.org/L2/L2007/07003r-n3194r-cyrillic.pdf

It certainly made sense to include this package in Unicode, and the vast majority of those characters certainly should be in this proposal. You do have to draw the line somewhere, and obviously those close to the line will be debatable, no matter where you chose to draw it, like this particular symbol - but once you've decided that you will include the one-eyed O (small and capital) and the two-eyed O (small and capital), then putting in the many-eyed O as well to complete the set doesn't seem so far-fetched.

Surprisingly many characters in Unicode are only recorded a few times if not once before the assignment. Chinese characters for example have a lot of them, because it was relatively frequent to make a new character for newborns before the modernity and some of them have survived through literatures but otherwise seen no uses (e.g. 𡸫 U+21E2B only appears once in the Records of the Three Kingdoms 三國志). But they have still received code points because they are considered essential for digitaization of historical works, and multiocular O is no different.

I didn't realize that digitization of all historical works was the goal of unicode. There's plenty of space for everything. And only a few fonts out there aim for complete coverage, like noto.

I just don't have the personal fortitude to attempt something so grandiose. Seems like a fool's errand.

Also, keep in mind there's not just one multiocular O. There's a bunch with varying numbers of eyes.

Not every goal needs to be something you can accomplish in a day or a year or even one lifetime.

There are not quite 8000 spoken languages on Earth at the moment, and a lot of them are from cultures that never invented writing. SIL has sent a missionary to most of them to learn the language, invent a writing system for it, teach it to them, and translate the New Testament into it. Most of those are fairly standard alphabets using characters from the Latin scripts, plus perhaps a few new characters or new combinations of character and diacritical. The task is large, but finite.

Imagine you’re a historian from the future studying some old document, and you spot a weird character that you’ve never seen before. Wouldn’t it be useful to be able to search for that character to see if it shows up in any other document? A simple OCR scan will bring up all the information you could ever need for that one weird symbol.

It's been seen once in the in-print wild.

There's no way to know how many since-written documents will break if a whole codepoint is dropped.

I agree with your mindset. It’s time for a unicode replacement.

Reply on news.ycombinator.com