Hacker News Reader

Comment by hoosieree | Hacker News Reader

> If Web 3.0 is already here, where is it, then? Mostly, it's hidden in the markup.

I feel like this is so obvious to point out that I must be missing something, but the whole article goes to heroic lengths to avoid... HTML. Is it because HTML is difficult and scary? Why invent a custom JSON format and a custom JSON-to-HTML compiler toolchain than just write HTML?

The semantics aren't hidden in the markup. The semantics are the markup.

View on HN

I think that’s what we’re doing today, and it’s a phenomenal mess.

The typical HTML page these days is horrifically bloated, and whilst it’s machine parsable, it’s often complicated to actually understand what’s what. It’s random nested divs and unified everything. All the way down.

But I do wonder if adding context to existing HTML might be better than a whole other JSON blob that’ll get out of sync fast.

I'm just not convinced that swapping out "<ol></ol>" for "[]" actually addresses any of the problems.

I must have missed your point, isn't the answer obviously that HTML is very, very limited and intended as a way to markup text? Semantic data is a way to go further and make machine-readable what actually is inside that text: recipes, places, people, posts, animals, etc, etc and all their various attributes and how they relate to each other.

Basically, what you are saying is already rdf/xml, except that devs don't like xml so json-ld came along as a man-machine-friendlier way to do rdf/xml.

There are also various microdata formats that allow you to annotate html in a way the machines can parse it as rdf. But that can be limited in some cases if you want to convery more metadata.

Why should anybody do that though? It doesn't benefit individual users, it benefits web scrapers mostly. Search bots are pretty sophisticated at parsing HTML so it isn't an issue there.

Web 1.0 = read

Web 2.0 = read/write

Web 3.0 = read/write/own

You could make the case that we already are in Web 3.0, or that we have regressed into Web 1.0 territory.

Back in actual Web 2.0, the internet was not dominated by large platforms, but more spread out by ppl hosting their own websites. Interaction was everywhere and the spirit resolved around "p2p exchange" (not technologically speaking).

Now, most traffic goes over large companies which own your data, tell you what to see and severely limit genuine exchange. Unless you count out the willingness of "content monkeys", that is.

What has changed? The internet has settled for a lowest-common denominator and moved away from a space of tech savy people (primarily via the arrival of smartphones). The WWW used to be the unowned land in the wild west, but has now been colonized by an empire from another world.

> Web3 = read/write/own

This is "reminagined" and more commonly known as just "Web3". Entirely different from the older conceptual "Web 3.0" in context of semantic web.

So more like:

Web 3.0 = read/write/describe

https://en.wikipedia.org/wiki/Semantic_Web

The resolution to this conondrum is of course to do both simultaneously and refer to it as "Web 5". Ask my friend Jack.

Reply on news.ycombinator.com