Hacker News Reader

The fact that in Windows-speak 'unicode' and 'UCS-2' (an encoding) have been used as synonyms has caused a universe of confusion. Even after I grokked the difference I still had arguments with colleagues who remained bamboozled by it. The fact that for a brief time there were only 64k code points, so that you could sort-of fudge the distinction, only made things worse in the long run.

View on HN

In my experience, most technical people these days still don't know the difference between UTF-8 and Unicode for example.

Is this the best reference? Too detailed?

https://www.joelonsoftware.com/2003/10/08/the-absolute-minim...

You found a quality link.

It does simplify things in places. One thing that isn't covered is that a code point is not necessarily a letter or vice versa. E.g. a single letter/symbol/emoji/whatever can be made up of multiple code points. So it isn't quite true that one "platonic" letter is one code point (it may or may not be).

If you deal with strings in Python 3 it kinda forces this distinction on you and I think is right to do so. It makes you painfully aware when you're assuming the world is utf-8.

Note that Windows uses UTF-16, not UCS-2.

It uses what's sometimes referred to as WTF-16. It's UTF-16 but with no actual guarantee that its well formed(i.e. invalid surrogate pairs).

Reply on news.ycombinator.com