Related ongoing thread:

HTML is all you need to make a website - https://news.ycombinator.com/item?id=33642490 - Nov 2022 (91 comments)

Don't miss a very probable context, that in the frontpage we currently have an older submission "HTML is all you need to make a website"

The bottom of the page already mentions that it's a direct response to that article

I've heard that you don't need JS neither CSS, now this is next level. Jokes apart, I think this is a readable page, despite the lack of centering.

It depends on what you are making, specifically.

I do think that usually you should not need JS nor CSS. Often HTML is not needed either and plain text will do (even, without Unicode), but HTML can help sometimes.

I think that Unicode is bad. Furthermore, Unicode does not have everything. EBCDIC is no good either, though. Also, the linked file still has HTML, although only a character encoding declaration and a <plaintext> command.

If you can't say it in plain text, it isn't worth saying.

This is still HTML though. Can't you serve plain unicode using the text/plain content type?

`content-type: text/html` but first line is just `<!doctype UNICODE><meta charset="UTF-8"><plaintext>` then no more tags, even to close.

It is impossible to close a <plaintext> element.

I thought this would be another weird new tag, but on the contrary:

<plaintext> is deprecated since HTML 2

https://developer.mozilla.org/en-US/docs/Web/HTML/Element/pl...

Absolutely, was just implying it was served as HTML but about as little as you can get, just a tiny preamble.

I'll eagerly await the ASCII-only "You Don't Need Unicode"

I disagree with the flowchart. Surely, since UTF-8 is backwards-compatible with ASCII, your writing should be encoded in ASCII if it was written by an American.

Well I see 6 HTML tag on the site

Pretty funny, though I would hesitate to call Salma Alam-Naylor's fruitless article "excellent"

It does raise some valid points: does color encoding belong into unicode? should we roll back bold and cursive alphabet unicode sections and relay these to fonts? or should we extend unicode to include horizontal and vertical mirroring of characters?

The bold and cursive Unicode have a "proper" use in display for mathematical symbols. You can't roll back these assignments anyway, and these are probably most dominantly used in math homework etc. produced in MS Word.

I thought Unicode already has mirrored characters?

There is a Unicode control code to change text direction[0] (arab etc). But no control codes to mirror individual characters or words vertically or horizontally. Nor are there general stand-alone mirrored letters, only a few select exceptions [1]

U+202E "OVERRIDE, RIGHT-TO-LEFT"

0x0F3B 3899 TIBETAN MARK GUG RTAGS GYAS ༻

0x0F3A 3898 TIBETAN MARK GUG RTAGS GYON ༺

And some funny trickery mapping roman letters to … whatever looks similar: https://fontvilla.com/mirror-text-generator/

Long ago there were pure ASCII text files with tutorials about computer graphics written by some guy from South Africa.

There were ASCII diagrams and everything :)

Damn it unicode: https://imgur.com/a/W1jQSZM

The whole thing is wonky as hell on mobile. The title looks somehow italicized, but in slightly different ways depending on the browser choice.

The author should have used a monospace font or a pre html element. Dang, they can't do that!

The opposite of sucks isn't rocks but rules, isn't it?

This was my (weak) attempt at referencing the venerable https://www.html5rocks.com/en/index

someone should make a javascript version of this

const newDiv = document.createElement("div");

const newContent = document.createTextNode("You Don't Need HTML!");

newDiv.appendChild(newContent);

And a webgl one.

Time to go back to BBS. Or teletext.

HTML can be easier to read though.

And more accessible too. I mean, I appreciate that this is intended as a joke, but imagine trying to use read-aloud software on this page.