Why not allow the user or client define the pagination experience?

You can get and read War and Peace (or other similarly large books -- I frequently reference Adam Smith's Wealth of Nations) at Project Gutenberg. As a single HTML file.

The ability to search for any particular text instance within that document through browser capabilities is useful.

As is, I agree at times, the ability to present the information as some sort of consistently paginated context.

I'm thinking that somewhere between the 21st and 42nd centuries, computers might advance to a state where this is in fact possible.

If I wanted to search a discussion thread I'd rather use a search function.

Why?

I suppose you're going to respond with something along the lines of "because it's better than the browser's search function".

But again: why is that?

Why do the built-in browser tools suck as badly as they do?

Example, outside of search: I realised some months back that a chief reason we have browser Reader Mode features, to uncrappify crappy site CSS, is because browser default styling ... is so horrible.

And again: why is that?

Why do we have craptastic CSS entity defaults when, with a very little work, every page could look, without any styling at all, pretty damned readable:

http://codepen.io/dredmorbius/full/KpMqqB/

(And if that's not your cuppa, then themes such that you get what you want -- Night Mode, or whateves.)

Whatever browsers are being built for, it's not the people actually using them.

/me looks suspiciously at Chrome => Google => Advertising....

Gee... I wonder why that might possibly be.

Well for one thing a browser search function can't distinguish between post content and the rest of the page. Also just one endless stream for a thread is practically unreadable.

And you realise that the browser has literally parsed the entier document object mode. If there's a semantic structure to be found, the browser knows what it is.

Again: why is cabapility-in-the-browser so impossible to provide?

Why isn't "search post" vs. "search comments" something scoped into a browser? If comments are a ... not entirely novel or unknown concept within Web design and activity, why are the abilities, say, to have browser-generated lists of authors, or search by author, or post/comment distinctions, or persistent reputations of commentators so that I can tell if @[email protected] is just an asshat or a space alien cat who really knows where his towel is (and independent of whether or not dredmorbius maps to any particular known "real name")?

It's not like the question's just come up within the past 5 hours.

Really: Why is the browser so fucking limited in this regard?

https://www.w3.org/TR/annotation-model/