HTML5 Review - My first attempt

Why would it even work – charset only applies to the tag it is declared on – you can even use it to MIX character encodings inside a document (what it’s for) – putting it on a meta does absolutely NOTHING in browsers that don’t know 5, and many that allegedly do!

Now, if it was on a tag that wrapped some content… like say… HTML? Of course that wouldn’t work either since in HTML 4 (and browsers designed for it) the ONLY TAG that has the charset attribute is ANCHOR… which is why on a standalone tag like META where prior to HTML 5 there is no charset attribute there is no way that should/would/could or even does work… So unless you’re setting it in your HTTP header, you’re not gonna have much success with it.

The entire preceeding two paragraphs to the statement and everything I just said above?

You can have zero h1s and use section, header, article and footer. Or only one. Or half. Or all. It’s up to you.

AGAIN, you missed what I’m saying. I AGREE with what you just said – that does not conflict or preclude what I’m saying. What I’m saying is that IF you use numbered headings and IF you use those tags, they all become h1’s because the count gets reset every time you open one of those new tags.

So where you would have in HTML 4


<h1>Topmost title of which all other headings are subsections</h1>

<div id="content">

	<div class="section">
	
		<h2>Start of section 1</h2>
		<p>Some text describing the subsection</p>
		
		<div class="subSection">
			<h3>Start of a subsection of section 1</h3>
			<p>Some Text</p>
		<!-- .subSection --></div>
	
		<div class="subSection">
			<h3>Start of a subsection of section 1</h3>
			<p>Some Text</p>
		<!-- .subSection --></div>
		
	<!-- .section --></div>		
	
	<div class="section">
	
		<h2>Start of section 2</h2>
		<p>Some text describing the subsection</p>
		
		<div class="subSection">
			<h3>Start of a subsection of section 2</h3>
			<p>Some Text</p>
		<!-- .subSection --></div>
	
		<div class="subSection">
			<h3>Start of a subsection of section 2</h3>
			<p>Some Text</p>
		<!-- .subSection --></div>
		
	<!-- .section --></div>		
	
<!-- #content --></div>

With HTML 5 you end up with


<h1>Topmost title of which all other headings are subsections</h1>

<div id="content">

	<section>
	
		<h1>Start of section 1</h1>
		<p>Some text describing the section</p>
		
		<article>
			<h1>Start of a subsection of section 1</h1>
			<p>Some Text</p>
		</article>
	
		<article>
			<h1>Start of a subsection of section 1</h1>
			<p>Some Text</p>
		</article>
		
	</section>
	
	<section>
	
		<h1>Start of section 2</h1>
		<p>Some text describing the section</p>
	
		<article>
			<h1>Start of a subsection of section 2</h1>
			<p>Some Text</p>
		</article>

		<article>
			<h1>Start of a subsection of section 2</h1>
			<p>Some Text</p>
		</article>
	
<!-- #content --></div>

I’m not saying you have to use them {endless string of frustrated expletives omitted} you {even nastier long string of expletives and personal insults about your family and level of education} :smiley: Joking – <ducky>it’s called a sense of humor, you should try one, it’s nice!</ducky> – I’m saying that if for semantic reasons you are using headings, the presence of heading, section, footer, and article reset them all into being H1’s instead of maintaining the cascade that actually made numbered headings make SENSE… NOT that anyone actually bothered using them properly which is probably WHY HTML 5 took that route… at which point ditch the numbered heading tags alltogether as they no longer serve a purpose; Just use HEADER instead. Header vs. Heading – redundant pointless BULL.

You’re missing the point, by the rules of HTML 5, using lower order headings like you are supposed to in HTML 4 is as broken/nonsensical as skipping heading numbers and having multiple h1’s under the old specs! You cater to the older UA’s by using lower order numbers, you’re telling HTML 5 structure where to stick it – you use the rules for HTML 5 structure, you’re telling older UA’s where to stick it… All because most people are too malfing stupid to understand the numbered heading tags in the first place! Something so simple a 2nd grader could probably grasp it – but with people these days graduating high school with the equivalent of a fourth grade education circa 1980, I really shouldn’t be shocked by this.

I really don’t think you’ve grasped what I’m saying, that or you are being deliberately obtuse on this… Are you joking and I missed the joke?

<header> is different to the <hx> tags, you can’t just use one instead of the other…

grabs popcorn

It works because people have historically used stuff like

<META HTTP-EQUIV=CONTENT-TYPE CONTENT=text/html; charset=windows-1251>

…and expect it to work in browsers, so they sucked it up long before HTML5.

No, <meta charset=…> and <meta http-equiv=“content-type” content=“text/html; charset=…”> applies to the whole page.

No you can’t.

It changes the character encoding of the document (unless there’s a BOM or HTTP content-type charset), that’s what it does.

Please leave your sense of logic at the door, thanks!

OK. The statements were wrong so the evidence has to be bogus. :slight_smile:

OK, then no need to discuss it further I guess.

I disagree. It’s allowed by the spec and makes perfect sense if you care about legacy UAs/ATs.

The HTML5 structure is identical (by design), so I’m sure it won’t mind.

I’m just trying to correct misconceptions.

Which again I don’t think people got my joke – That was a “oh wait, too late”; an indication of the idiotic levels the “specification” has reached. I was being flippant, indicating they had already done so. (hence the word facetious?) When a specification for something as simple as HTML is big enough to warrant SVN, it may just be time to pitch it in the trash and start over.

Interesting history of HTML specs though, as formatted web pages:

HTML 2 - 11k (flat)
HTML 3.2 - 122k (+32k of images)
HTML 4 - 53k (+3k images + 3k CSS)
HTML 5 - 4,824K (+601k images +2k CSS)

Oh yeah, some real “improvements” there. When a specification goes from the size of the Declaration of Independance to larger than the KJV with a ribbon bookmark, it MAY be time to consider it a wee bit overcooked.

Admittedly, MUCH of that is all the NON-HTML stuff they’ve thrown under the HTML 5 banner – but still, we’re supposed to be taking web development queue’s from a group that publishes the new standard as a single 5 meg website? Much less it was bad enough the endless legalese the earlier specs were written in – Without the markup as flat plaintext it’s still 2.3 megabytes; The equivalent of ten sitepoint books or four James A. Michener novels… NOBODY except total nutjobs are going to actually read that as a technical specification!!! Lands sake the technical manuals for a F-16 are smaller COMBINED… and trust me, a Viper (Falcon for you gropo’s, “Fighting Falcon” for the civvie pukes) is and should be significantly more complex than something as moronically simple as HTML should be.

Probably why every time someone says “uh-uh” or talks about HTML 5, I can point them at where the spec says “uhm, no…” or at the very least does the exact opposite in their examples… I’m one of the few crazy {string of expletives again omitted} far enough out there to even ATTEMPT reading it end to end.

Which I have… and of course by the time I finished it a hefty part of the content I’d read had already changed. The problem with using anything that’s still in DRAFT.

The global structure of an HTML document

Is not, has not, nor ever has been part of the HTML 4 or 3.2 specification as a valid attribute on META. The only place that ever existed as a fully working element was Netscape 4, which by extension Gecko picked up… IE kind-of mimicked it, but doesn’t set the value properly in the DOM completely shtupping scripts who will do things like read innerHTML as the wrong encoding.

Does NOT work in Opera, Does NOT work in Webkit, Does NOT work properly in IE (though it tries)… Only place it works is in Gecko based browsers, which makes sense them being Netscape 4’s sweetly retarded cousin and all.

It’s a perfect example of undoing everything HTML 4 STRICT was about… like say… using the actual HTTP commands to do ****. You know in theory ANY response header element should be able to be sent that way? Of course browser makers dragged their heels on actually doing it – so much like Object they’ll throw it in the trash for some other technique that they’ll never agree on implementing the same way.

All though this is a nice attempt. You should keep in mind that you have to continue this learning. Otherwise this will bear nothing.

That’s not relevant.

Please show me your test case where it doesn’t work.

Part of the problem of the older specs is that they were too ambiguous and left a lot of things open to interpretation of the browser makers, thus giving us the mess where you have several browsers rendering the same things in different ways. Of course, measuring how good a specification is by the filesize of the document that describes it is some pretty stupid logic right there, but I only see more specification for edge cases as a good thing.

Also, FYI, CVS/Subversion has no minimum size of project that it can cater for, and proves just as useful for a single document as it does a whole project.

You mean the entire POINT of HTML? Device Neutrality? The reason for strict being to move all the ambiguous stuff out of the markup in the first place?

Putting the blame ENTIRELY on how browsers allow elements to receive CSS instead?

You mean the original POINT of HTML 1, 2 and 4 strict – the thing people got away from with the disaster called 3.2 that it seems HTML 5 wants to bring back? The whole idea being to say what things are and let the user agent best determine how to show them – or with 4 strict using CSS to tell the user agent (browser for you normals) how to show it?

JUST talking markup we’re talking 115 tags and around 40 valid attributes… As such we’re talking 14k of text per element – that’s BULL. Especially since as HTML should have NOTHING to do with appearance all you need to do is say “this tag has these attributes and means this”. When you need twice the size of the Declaration of Independance to describe a HTML tag, you need to go back and learn how to write. When your specification for something as simple as HTML is three times the combined total technical documentation of a US Military fighter jet, and 100 times the size of the US Constitution WITH the amendments…

But then, I used to write technical manuals for Marstek – where I learned the rule “if it takes you more than a page to explain it, you’re explaining it wrong”.

Which is just part of why I say HTML 5 is crafted for the people still vomiting up 3.2/4 tranny and never grasped the point of strict, the advantages of separation of presentation from content, etc, etc…

If you even use the word “rendering” in regards to HTML, you have missed the POINT of HTML 1, 2 and HTML 4 strict and are basically still in “transition” from 1997 to 1998… and you’re laying the blame on the wrong part of the specification’s lap. (the problem being elements like tables and inputs do not receive CSS the same way, browsers don’t render named border styles the same way, etc, etc… that’s not HTML’s fault!)

Well no, REALLY? Completely missing my meaning in that it shouldn’t be complex enough to NEED the software assist. That it is ends up just another indication of what a bloated useless train wreck NOBODY is going to take the time to understand.

… and to think the framers of HTML 4 called 3.2 “uselessly large”.

Indeed. Part of the problem with HTML5 is that even the creators of it don’t actually seem to understand it. Like what <nav>, <section>, <article>, <header>, <footer> and <aside> are actually for. (Everyone seems to have a different opinion about it, but it shouldn’t be a matter of opinion.) It’s no wonder everyone else is so confused. :frowning:

Well they obviously failed to grasp the point of STRICT or concepts like separation of presentation from content – why should their comprehension of their own nonsense be any different?

Of course when it’s over 2 megabytes when converted to plaintext, who the devil is even going to remember any of it, much less comprehend it or even take the time to read it?

As I said though, rather a lot of that is instructions to browser makers on how to handle different elements, unambiguously, not everything in there is needed to USE the language

Which gets away from the concept and point of HTML 1, 2 and 4 Strict, and sounds a HELL of a lot like HTML 3.2 – Since the default handling of HTML elements without explicit instructions from CSS should be ENTIRELY at the discretion of the user agent. If you’re going to dictate how the elements work without CSS to the user agents, that SOUNDS like a return to presentational markup and all the crap STRICT was trying to get rid of. (and part of why XHTML 1.1 was doomed to failure as it too had that same shortcoming and philosophical failing – trying to turn markup into something it’s really not meant for)

It’s only when stylesheets are on the table that said behaviors should matter… but that’s part of my problem with HTML 5… it’s not JUST covering markup, they’re slapping every other blasted specification out there under it’s banner – which of course is where all the useful stuff actually is. CSS3, the new scripting elements, aria, etc, etc, etc… Basically all the stuff that has NOTHING to do with markup.

Which is what HTML is SUPPOSED to be about. Of course, if you took all that stuff away from HTML 5 it would be a rather pathetic, puny and pointless thing.

Let’s face it, all the really cool “HTML 5” stuff has nothing to do with HTML – CSS3, the new scripting controls and scripting related elements (which is why for example I say canvas shouldn’t even exist as a tag and should be added by scripting not markup), etc, etc… That’s where the cool and useful stuff is.

Since the default handling of HTML elements without explicit instructions from CSS should be ENTIRELY at the discretion of the user agent

What a load of rubbish! The visual rendering of them maybe, but not how to handle the elements and how they should function.

Sorry, but I think you’re allowing your anti-5 prejudice to get in the way of rationality here…

The tight specifications are around default presentation, behaviour, role and meaning of each element. They are not about how to get the visual presentation that you want using HTML, they are very clear that that is what CSS is for (with the exception of the blasted <font> tag). I don’t see how a system that gives consistent implementation across different browsers can be a bad thing.

I see not the distinction between default presentation or behavior, and presentation as a whole. The first two you list have no business in a HTML specification if working to the original intent.

… and frankly if it takes more than an average 5k of plaintext to explain the element in a specification format for BOTH browser makers and the people shock actually expected to use this shlock, there is something HORRIBLY WRONG with what’s being implemented; ESPECIALLY with 99.99% of tags sharing the same grammar/syntax. The biggest distinction between them being unique non-common attributes, is it inline-level or block-level, what can it contain, what can it be contained in, as inline can it accept block parameters (img and textarea for example), and what is the tag FOR semantically.

Anything beyond that has no business in a markup specification where we are allegedly supposed to be saying what things are, not how they appear.

… and would give you around 500k total, not 2.3 megabytes. Of course if like tags were actually combined down and some sensible content management were implemented, someone kicked whoever it is that has the love for cryptically vague legalese where it really hurts… it would probably be even less than that.

Dude, you’re free to fork the spec and throw out everything you think doesn’t belong in “HTML”. As someone who works for a browser vendor, I know everything is needed to be defined (I don’t care if it’s in a spec called “HTML” or if it’s split across several specs). Here’s one fork you might be interested in, although I’m sure it’s still way to much detail for you to be satisfied. HTML5 — Edition for Web Developers

Sorry if i don’t get the point here. But what does images and css have to do with html5? I mean the images and css are there for the presentation. More images because nobody wants a site that looks like it was looking before 15 years (expect deathshadow60 :slight_smile: )

OBVIOUSLY… You completely missed the entire point of the section you quoted.

Those are the sizes of the W3C pages FOR THOSE SPECIFICATIONS, as in how big the pages are… In other words HTML 4 required a tenth the images to explain itself that HTML 3.2 did, while HTML 5 requires 200 times as much. The initial numbers is how many K of markup (html) those specifications take – meaning that from HTML 4 to HTML 5 they’ve grown the spec to over 90 times the size.

Though that’s unfair – with HTML 4 they had the common sense to break it into separate pages and I’m only listing the main page – instead of HTML 5 where they’ve dumped everything into one massive document that brings the fastest browsers on bleeding edge machines to their knees.

…and the fact that the spec is all on one page means the whole language is rubbish how exactly?