Guest post by Fatal Exception’s Neil McAllister questions whether the ideal of a truly standards-compliant Web will ever be possible given ongoing disagreements over HTML 5, including those most recently set in motion by Microsoft’s latest concerns over the spec. “While some items on Microsoft’s list seem like quibbles, others are valid issues,” McAllister writes. And with others in the working group supporting the proposed additions in question as they stand, an impasse could conceivably be the end result. “Disagreements are an inevitable part of any standardization process. But if disagreements lead to irreconcilable differences, it calls in to question the validity of the final standard.” So where does that leave HTML 5?
Let me see, perhaps they want the elements to more clearly sync up with traditional publications and stop using class and id to create these element intentions?
Strict readability of code is an improvement.
I’d imagine latex2html5 would appreciate elements as straightly translatable as possible.
Edited 2009-08-28 20:34 UTC
I agree with that. Especially for those who code HTML “manually”, it would be an improvement if they could describe what some contextual element is, instead of describing how it should look like in the first place. Of course, the div and span tag allow to do so, but from my own experience in “HTML vs. LaTeX”, I think the structural (vs. the visual) way is more comfortable. Not for everyone in every place, I’m sure, but still a promising approach.
The whole point of HTML5 was that neither Microsoft nor the W3C was creating what DEVELOPERS and Browser makers needed so they formed WHATAG.
Things like header, footer, and the various formatting identifiers are important to screen readers and AJAX scripts reading web pages for other purposes. This lets them make assumptions about what’s content and what’s fluff. Things like “mark” take advantage that not all web browsers are dumb clients, some like Amaya allow reading and writing at the same time… we need more of that!!!
The people that developed HTML5 have spent years actually writing HTML pages! They’ve been doing it since before the 4.01 spec was final and they want shortcuts in the markup to follow the standard way pages are crafted after nearly 10 years of non-development.
The Microsoft guys are playing the typical game played in committees of showing up with big names, showing up late and unprepared, and “we didn’t used to do it that way”… the point being not to construct solutions but simply to say they were at the meeting while not adding any more work to their plate.
My advice is to lock MS out of the talks, and let them find out on the ‘tubes when it’s done.
The only reason why IE8 is more standard compliant today IMHO is because Firefox got so popular that web developers started to complain to MS for having their own standard. It doesn’t matter why Firefox got so popular (IE6 was too vulnerable, Firefox allowed extensions, performance reasons) it just matters when your website viewers/visitors swap browsers that you can actually still support them and having 2 standards make life difficult with developers.
If companies could be fined for using non standard code on their www websites then it would force them to actually use standards. Same goes for browsers. This argument ties into net neutrality IMHO.
Someone please tell my why is the PS3 browser blocked from Hulu.com, or why http://digitalmaps.com is policing me into using internet explorer even though I have version 8 and the site doesn’t work with version 8 (tried it on 3 computers all with IE8). This frustration is making me want to buy a rifle and climb a tower.
Edited 2009-08-28 22:47 UTC
There is another way, but I’m sure no browser will go for it: Imagine that the browser contains a check for valid HTML. It’s always possible to specify which HTML you are using (DTD), so it searches for the specification and checks if it is compliant to the respective standard. If it is, it displays the page. If it is not, it does not display the page. Displaying an error message (“The page you requested is not a valid web page.”) or show a yes/no dialog (“The page you requested does not contain valid content. It possibly cannot be displayed correctly due to mistakes the developers of the web page made, Do you want to try to display it?”).
You’re mentioning “net neutrality”. Let me be more precise: Client neutrality. You should be able to view content, no matter which client you use, especially when there’s no reason that content should not be accessible. Of course, nobody expects to be able to view images with a text mode browser, or “Flash” movies with an FTP client. But I’m sure you get the idea. It’s about accessibility. Remember, there are blind users out there who are happy when someone uses alt= and longdesc= for the img tag instead of stuffing the whole navigation into “Flash”, and designs his web stuff properly so that it can even be read by a speech synthesizer.
See above: Mistake the developers made. If someone makes a web page that is not a web page (at least not a standard one), then the publisher does not want you (with your standard compliant browser) to look at it. Why should you excessively try to look at something that someone else won’t let you?
Instead, buy a tool that lets you arbitrarily delete such stuff from the web.
Wouldn’t that be lovely?
SGML has had that from the beginning. HTML is loosely based on SGML … but left all the good stuff behind. XML was supposed to bring things back to the glory days of SGML and valid documents … but was then perverted into XHTML which doesn’t require valid markup (at least, not always).
What’s the whole point of having the DTD tag/info in HTML pages … if it’s not used to validate the markup and the data?
XHTML is not a perverted form of XML at all. XHTML documents are required to follow XML rules for valid markup. 99.9% of the time XHTML is sent as text/html as IE does not support it, as such it is treated as HTML tag soup by browsers and so using it over HTML 4 is pointless. (the doctypes are identical sans the xml namespace attributes)
That only leaves, do browsers validate against the schema/doctype, which possibly they should, don’t think they do. This is also a big task as being XML they could pull in any schema type (think MathML, SVG etc.) which the browser may not know of, and the namespace uri does not have to point to a copy of the schema.
For example, <audio> and <video> are completely useless.
What was wrong with the old <object> tag?
I’m no HTML author/”programmer”, but I also read pretty harsh comments about the new <div item> tag.
Like you said, you’re not web developer.
The audio/video tags encapsulate all the ugly that object does not. Flash comes to mind as how damn ugly it is to get it to work and all the hacks here and there for CDATA and earlier browsers.
So new elements <audio> and <video> are needed to fix the old unsatisfactory handling of <object type=mimetype> by clients?
That doesn’t exactly make sense… <object type=mimetype> means to the client: “Here is an object of type mimetype, can you handle it?” It’s then up to the client to use some way to display the content. If the client uses some shoddy plugin, that offers incomplete control of the content, it’s not HTMLs domain anymore.
One feature of <audio>, <video> are script accessibility and playback control. While this is nice for “multimedia application” websites, I feel it’s wrong. The document written in Hypertext Markup got served… not the application written in Hyperapplication Markup, or some such.
What the W3C, WHATWG and their ilk should be doing, is separating applications and documents and the way each is served. So they could create HAML (Hyper Application Markup Language) which is transfered via HATP (Hyper Application Transfer Protocol) and deprecate all scripting in HTML to turn those back into the static interlinkable documents they once were.
Then those “web designers” that need to make every document as flashy, blinky, clicky as possible, can just use HAML and they and the W3C can bash each others heads in over (scriptable) <audio>, <video>, (the clearly application related) <progress> and such.