Text encoding of spec.html: ascii or utf-8?

[claudepache]

Currently, the file spec.html contains the declaration

<meta charset="ascii">

which is contradicted by its contents, because it has two non-ascii, utf-8-encoded characters (namely ≥ on line 7524 and ’ on line 38018).

That makes my text editor complain each time I want to edit the spec, either:

or:

Two options:

1. Officially allow utf-8-encoded contents, and declare formally spec.html’s charset as utf-8;
2. Replace the two offending characters by the proper html entities, and add an automated test in order to reject further insertions of non-ascii characters.

Personally, I prefer option 1 (≥ is more readable, more writable (although I guess that may depend on your working environment), and I never recall the html entity for that character).

[jmdyck]

The U+2019 right single quotation mark on line 38018 should be a U+0027 apostrophe. Commit 64ab8cf in my "Misc editorial" PR fixes this.

So it's really just the U+2265 greater-than or equal to on line 7524. For option 2, this could be changed to ≥.

[annevk]

Note that <meta charset=ascii> isn't really conforming HTML anymore (and conforming implementations will treat it as if you wrote windows-1252). Only UTF-8 is allowed.

[jmdyck]

The 'rendering' process generates a file with <meta charset="utf-8">, so at least we're okay on that front.

[claudepache]

Resolved by dc1e21c, which follows from the introduction of character ℝ in spec.html.