Re: [I18ndir] [art] New Version Notification for draft-bray-unichars-06.txt

[Rob Sayre <[e-mail]>]

On Mon, Oct 9, 2023 at 4:38 PM Martin J. Dürst <[e-mail]>
wrote:

> Hello Rob, others,
>
> On 2023-10-03 01:42, Rob Sayre wrote:
>
> > Basically, UTF-8 is a work of genius, but it arrived 10 years too late.
> > JavaScript, Windows, Java, and all the rest of them were underway before
> we
> > got RFC 2277. Things would be much different if it had appeared in 1982
> and
> > everyone then realized that you don't need Θ(1) access to characters in a
> > string...
>
> You make it sound like Java already existed in the (early!) 1980ies, but


Well, I know how to write Common Lisp and C++ pre-98, but I was not trying
to take the conversation in that direction. I have a 2005 Powerbook right
here, and I have managed to connect it to the internet, but it is a bit
rough.


> the history is much more compressed. Java started in the early 1990ies.
> And RFC 2277 was published in 1998, but before that, there was RFC 2130,
> and the workshop that was the base for it in Feb/March 1996.
>
> Another way to express the problem is to say that the emoji craze should
> have happened much earlier, because emoji beyond the BMP (the original
> 16-bit space) were the main driver to get UTF-16 taken seriously in
> programming languages such as Java and in databases :-).
>

I agree with your point here, but I did write something close a few months
ago. I cited this rationale:
<https://simonsapin.github.io/wtf-8/#motivation>

It cites emoji, if you give it a close read.

UTF-8 was around, but the web was not in UTF-8 then, so it was still
dealing in many other charsets, and the operating system strings were still
uniformly awful. This comment does not mean I am advocating sending WTF-8.
But that data format still underlies several different systems, so it
arises in escape sequences etc. This draft is about repertoires of code
points, so we still have to describe it this way, if only to exclude them.

This document describes a 2020 rewrite of charset detection:
<https://hsivonen.fi/chardetng/>

I am in favor of describing the bad parts, but I do not view it as an
endorsement.

thanks,
Rob