So I’ve been making HTML pages for some time now and got to this conclusion that HTML itself has a lot of really useless things. And I don’t understand intensions of creating them. Like, we have lots of meta tags, such as “keywords”, “author”, “generator”... They are never used to show information on the page, so it was easier to just write that data as comments since you have to see source code anyway to know what’s in “author” and “generator”...

But even more useless to me look semantic tags. They can’t express everything but when they do express something — you normally won’t see it. For instance, we have “var” tag for variables and that’s italics by default. What’s the point of it? If you need to make something italics, just make it so. And if you try to use every sematic tag out there correctly — noone will ever see your diligence. Or if you want it you should probably say in the beginning of a page that reader better read source code.

And still you don’t have semantic tag for everything. We have “var”, right? Where are tags for classes, functions and methods?

Those are for search engines to read.

SE don’t buy semantics anymore.

If you have marked variables separately then you can changes the formatting of just the variables with CSS. In fact, some browsers allow you to apply custom CSS to webpages so everyone could have their own formatting scheme. Of course, since the not everyone uses semantic markup this doesn't work in practice, but you can still see the remains of the Grand Vision in the HTML spec...

They realized (after already having defined "var") using the class attribute was a more extensible way than defining every possible semantic tag that anyone might ever want, I guess?

Hi OP,

var tags are part of the java script scripting language. It is just plain text. I think that it is your IDE or view source output of your browser which is just showing the word in italics.

As to meta tags, they are used to show information to the search engines.

OK

It's not JS, it's HTML.

Again - not any more. SE just don't care for them these days.

Remember that HTML is a simplification of SGML, which is used for marking up documents; Sir Tim Berners-Lee encountered that at CERN. The whole point of a mark-up language is that it marks function, not format. In marking up a MS for a book, you'd use something like <sq> of short quotations and <lq> for long ones. The editor can then decide whether <sq> will be realised as single or double quotes, and whether <lq> will be the same or an indented paragraph in small type. Tags like <cite> enable automatic creation of bibliographies. That sort of thing is still useful in HTML, because it's not confined to the web: it can also be found in databases and internal documentation.