Have you ever looked at or tried to implement an HTML parser? The language creators idiotically decided to allow unclosed tags such as <br> and <input>, which means that you can't just watch for the end tag while parsing but have to keep a list of which tags are allowed to be unclosed and which aren't. They also decided to make it part of the standard that mistakes such as <p>...<div>...</p>...</div> should be automatically reordered by the parser into <p>...</p><div>...</div>. An HMTL parser SHOULD be 100 lines of code long or so. Instead one is typically several thousand lines long. Dumb dumb dumb.
Personally, I also don't like tag names such as "p", "div", and so on. These define semantics and behavior, but the idea of semantic HTML is absurd nonsense. As for the behavior, why not put it in an attribute. Instead of:
<div class="someclass" id="3">...</div>
I'd rather write:
<tag="div" class="someclass" id="3">...</div>
This would make a parser even easier to write and be more consistent. HOWEVER, I understand that they wanted HTML to be a derivative of XML, so I'll let this one slide.
Here's another one. Browser layout is GARBAGE. That almost goes without saying, of course. The sins of browser layout are well-known, although personally I feel that it still does not get the treatment it deserves. In an ideal world. its designers would be excoriated and widely MOCKED for this garbage. Flexbox is only now becoming widely adopted. You fucking kidding me? Its capabilities should have been part of browser layout from DAY ONE MOTHERFUCKERS.
Here's another one. REST is GARBAGE. HTTP verbs are a DUMB idea idea. Get/Put/Post/Delete? Fuck you! There should be only one type of HTTP request. There is no way to absolutely or ideally classify request semantics, so this should be left up to users.
>Hasn't heard of XHTML and W3C's subsequent cucking to WHATWG
>Hasn't heard of HTML5 semantic tags
Would a cock in your mouth calm you down, sweetie?
>>61347072
>Hasn't heard of XHTML and W3C's subsequent cucking to WHATWG
Tell me more.
>Hasn't heard of HTML5 semantic tags
You mean <form> and <table>? Of course I've heard of them, but it's a bad idea. Semantic stuff should be defined in attributes.
>Would a cock in your mouth calm you down, sweetie?
Probably not.
>>61347260
>You mean <form> and <table>?
HTML*5* ones: https://www.w3schools.com/html/html5_semantic_elements.asp
>>61347277
Still a dumb idea to bake it into a standard. Any such classification of page components will be insufficient and break. You have to let the web developers build custom elements themselves, not try to classify elements for them in advance.
>>61347001
>"p", "div"
you mean paragraph and division?
>>61347299
>Markup amenable to parsers
>Freedom to into your own whacky shit
Pick one and only one.
>>61347306
Yes. I think these are dumb. They might as well be class names. What's the point of having them in a special super-attribute of the element called a tag name?
Of course, some tag names such as "a" and "input" define behavior. But I'd still prefer that these be defined by simple attributes. For example, instead of:
<a href="www.porn.com">Porn!</a>
Do:
<behavior="link" href="www.porn.com">Porn</>
>>61347334
But I'm not talking about making new wacky shit, I'm just saying that the idea of adding "article", "section", and other such "page component" types of tags to HTML seems pointless to me because it adds a bunch of arbitrary stuff to a language, whereas ideally the language should have as little arbitrary stuff as possible.
>>61347362
I believe the web as we know it with html/css/js is meant to be replaced someday as it is a fucking mess but what I have no idea is how the transition will be
>>61347416
I think that to make a switch happen, you would have to write a new browser that runs a better set of technologies, but it would be able to also run standard html/css/js websites with performance as good as or almost as good as the best modern browsers. So it would auto-detect old-style sites and parse them through, say, Chromium... but for new-style sites it would use its own engine.
>>61347416
Best I can figure is a game-ified WebAssembly VR neurostim, but my imagination is limited by shitty cyberpunk sci-fi.
>>61347362
Write your web pages in XML instead. The major web browsers are happy to parse XML and apply CSS or XSLT stylesheets client side, so then you can use whatever elements you want.