Apple, Opera, and Mozilla Push For HTML5

foo fighter writes "The World Wide Web Consortium (W3C) has been slumbering the past several years: HTML was last updated in 1999, XHTML was last updated in 2002, and no one is taking seriously their largely incompatible work on 'next-generation' XHTML or 'modularized' XHTML. Both HTML and XHTML are in sorry need of removing deprecated items while being updated to reflect the current practices of web and browser developers and remaining compatible with legacy Recommendations. The much more open and transparent WHATWG (Web Hypertext Application Technology Working Group), formed in 2004 to address this problem, and has been hard at work on developing a draft spec for HTML5 to update and replace legacy versions of both HTML and XHTML. The quality of this work has reached the point that Apple, Opera, and Mozilla have requested the adoption of HTML5 as the new 'W3C Recommendation' for Web development."

Comments

The More they add, the less I like

[ackthpt]

OK, I'm a curmudgeon. There, happy?

I still design pages using HTML 3.2 standard. Life was happy when pages were small and simple. I'm very put-off by the way HTML now can do things formerly reserved for javascript. Further, people no longer appear interested in the size of the footprint their pages make and the bandwidth necessary to download them.

We rail away at Microsoft and anyone else who adds bloat to software, but the web is now plagued by page bloat and overly clever designs which render poorly at times, take over the browser and sometimes crash it. Behaviour is becomming terrible, but as pages are done by authors who do not really care, so long as it looks like it should and does the basics, they care not what a wreck have created.

Don't even get me started on people whose home page is some massive flash object.

"Hi, we assume you have the latest browser and all the plugins!"

Re:The More they add, the less I like

[KenAndCorey]

I'm with you brother, although I use HTML 4 and CSS 2. I wish people would take the time to code their pages so they are fast loading and elegant (code-wise), and HTML generation apps would do likewise. Additionally, I wish people would use proper caching as well -- this really speeds a site up too.

Re:The More they add, the less I like

[Bogtha]

I actually find things like "normal <B>bold <I>bold italic</B> italic</I> normal" useful

I hate to break it to you, but that's not HTML 4.01 Transitional either. No version of HTML has permitted overlapping elements in the way that you describe. You are merely exploiting error handling that is fairly common amongst web browsers.

Re:The More they add, the less I like

[Billosaur]

Yes, this incessant pushing of the technology/standards envelope is creating a lot of disjoint, stilted, and otherwise unreadable web sites. It used to be web pages were mainly HTML with a few SSI thrown in for good measure; now they are over-burdened with flashy graphics, tricky menus (god how rollovers are getting out of hand!), and a lack of decent content. I mean, I go to a web site to find information I'm looking for. In the old days, you could do that -- now content is so snarled in meaningless fluff that have the time I have to search the source code just to find what I'm looking for.

Re:The More they add, the less I like

[ackthpt]

I'm with you brother, although I use HTML 4 and CSS 2. I wish people would take the time to code their pages so they are fast loading and elegant (code-wise), and HTML generation apps would do likewise. Additionally, I wish people would use proper caching as well -- this really speeds a site up too.

I haven't done much with style sheets, finding them to be just one more thing to manage, as they can get rather large the more I relied upon them.

Effectively when we write the HTML code by hand we're creating very lightweight pages. I set some colours and a simple background based upon a small sample and I'm good. I came from the K-I-S-S school of web design, which seems to be dying mostly thanks to webapp/webpage development tools. It's like watching people program without a care about optimizing for size or speed. They're paid by the hour, not for the quality of the code.

Re:The More they add, the less I like

[mstahl]

Hi there. I'm a web developer/designer. I do flash, too. Good times, right?

I design and build to the XHTML 1.0 transitional standard, and for some bizarre reason one of my clients still makes me test their pages in IE5. When was the last time you even saw a computer that had IE5 on it?

Your objections to design I can't really comment on beyond saying I hope you're not referring to any of mine. But your objection to HTML/CSS doing what javascript used to be necessary for? Really? You prefer writing little-stupid javascript functions to just putting a :hover rule in your CSS? Really?

You, sir, are a rare breed. Hats off to you though; HTML 3.2 is really the only standard the most browsers agree upon (IE6/7 have all those weird box model problems with XHTML 1.0).

Re:The More they add, the less I like

[LighterShadeOfBlack]

Life was happy when pages were small and simple. The Internet was also small and simple, relatively speaking. Unfortunately now it's a huge mess of information, some useful, some not. In order to helpfully and meaningfully wade through all this fluff we need to more tags and more specificty in our markup to aid search engines and the like in finding what we really want. We may be a way off from the "Semantic Web" as Berners-Lee envisions it, but these are the first steps towards making that happen and preventing the web from being collapsing under it's own ever-increasing mass.

I'm very put-off by the way HTML now can do things formerly reserved for javascript Yeah, that never happened in the past <blink>Remember me?</blink>. Seriously though, I agree on this in principal although I'm not sure specifically what features in HTML you're referring to. Ultimately any attempt to dynamicise (I know, I know, not a word) HTML will fail as it will always be three steps behind what people want from dynamic web pages since we're now moving into the whole "Web 2.0" thing.

Further, people no longer appear interested in the size of the footprint their pages make and the bandwidth necessary to download them. I'm not sure I agree with this. Relatively modern developments allow far more efficient web pages. Firstly by using CSS you can do a lot more with simple markup while allowing the stylesheet itself to be cached for a reasonable amount of time (whereas many webpages have content which prevents long-term caching). XmlHttpRequest obviously allows for only the relevant portions of a website to be updated. Javascript allows for less data to be sent and for the code to do the work of constructing an elaborate webpage (only applies to certain types of webpages obviously).

We rail away at Microsoft and anyone else who adds bloat to software, but the web is now plagued by page bloat and overly clever designs which render poorly at times, take over the browser and sometimes crash it.

...

Don't even get me started on people whose home page is some massive flash object. Sure, some people use poor designs which drain resources unnecesarily, I don't think that's necessarily an issue of new standards or technologies being poor though, just that the flexibility we demand from our new web technologies inevitably allows for misuse. You can't blame Javascript, XHTML, or even Flash simply because some people will misuse it any more than you can blame HTML 3.2 because someone decides to use 24 levels of <table><tr><td> tags to make their layout the way they want. As far as crashing goes, that's a software issue and nothing more.

Re:The More they add, the less I like

[Doctor-Optimal]

They can have my blink tag when they pry it from my cold, dead hands!

Re:The More they add, the less I like

[Bogtha]

It's like watching people program without a care about optimizing for size or speed. They're paid by the hour, not for the quality of the code.

Funny, that's how I feel about people who don't use CSS. Seriously, if you are that concerned with the size of pages and bandwidth, like you say in your other comment, then why are you transmitting your style information on every single page load?

Re:The More they add, the less I like

[nine-times]

I don't really get your complaint. I mean, I share your annoyance with uselessly flashy pages, and literally Flash-y pages, but what's wrong with refining standards? Many of the updates to HTML have made things cleaner, more precise, and more consistent. Some of the added features have allowed web developers to do more with less code (if you can call HTML "code"). Much of what's added in-- if you don't want to use it, don't use it. But if you have some reason to do something flashy on your site, it's probably better to have it be done in some standard way rather than though some hack or by adding Flash to your page.

Re:The More they add, the less I like

[Chris Mattern]

They can have my blink tag when they pry it from my cold, dead hands!

That...can be arranged.

CHris Mattern

Re:The More they add, the less I like

[Bogtha]

You prefer writing little-stupid javascript functions to just putting a :hover rule in your CSS?

I get the impression he's not a professional web designer, so he can just ignore stuff like that entirely.

HTML 3.2 is really the only standard the most browsers agree upon

There's a very good reason for that. The W3C were working on HTML 3 when it became apparent that their work was diverging from what browsers understood; browser vendors were adding stuff at a crazy rate while ignoring the HTML 3 work. So the W3C decided to scrap HTML 3 and make a decent description of what browsers understood in HTML 3.2.

Basically, the reason why "most browsers agree upon HTML 3.2" is because HTML 3.2 was merely rubber-stamping what browsers already did.

IE6/7 have all those weird box model problems with XHTML 1.0

There's no such thing as a "box model" in XHTML 1.0. The box model is a feature of CSS.

Re:The More they add, the less I like

[blincoln]

Funny, that's how I feel about people who don't use CSS. Seriously, if you are that concerned with the size of pages and bandwidth, like you say in your other comment, then why are you transmitting your style information on every single page load?

Agreed.

To the GP: I recently redesigned my main website after running it for five years with a design very much like the one you describe - all coded by hand, HTML 3.2, no CSS (although I had some equally old Javascript for highlighting the navigation buttons).

The new version uses CSS, and since I designed it using the "strict" mode of newfangled HTML, it renders more or less identically on different browsers. I also built a batch build content management system, so that I don't have to manually edit a bunch of HTML when I change the design or whatever. I made sure the output is basically what I would have done if I did it all by hand though.

I was very skeptical about it before I started, but it really is a much better way to build websites. It saves time, it makes redesigns and multi-platform stuff easier (like theoretically I could swap out CSS files to make a version formatted for PDAs if I were running a website that would be at all useful on them), and it's *much* easier to get relatively consistent rendering across platforms. The only visible difference I'm aware of between Firefox and IE6/7 is related to tables without a fixed width. Neither one looks superior, they're just different.

Re:The More they add, the less I like

[Bogtha]

Yes, because MS FrontPage would never output broken HTML, would it?

Re:The More they add, the less I like

[EsbenMoseHansen]

Effectively when we write the HTML code by hand we're creating very lightweight pages. I set some colours and a simple background based upon a small sample and I'm good. I came from the K-I-S-S school of web design, which seems to be dying mostly thanks to webapp/webpage development tools. It's like watching people program without a care about optimizing for size or speed. They're paid by the hour, not for the quality of the code.

Why do you set color and background? Let that be up to the user. Structured text with some hyperlinks, that is the way to go! Emphasize text with the em tag, use h1-h3 for headers, and the list tags... maybe the table tag for a simple table, or if extreme vital, a link to a SVG or PNG image (which should be obvious). That should be all you need. If you need personal, do a favicon, or maybe a link to a personal picture.

;)

Re:The More they add, the less I like

[Nimey]

What's a French?

Re:The More they add, the less I like

[Anonymous Coward]

Maybe I'm missing something very important but using HTML 5 to describe HTML 5 seems _stupid_ to me. Like learning an adult French using only French. Obviously, it was also stupid to think that a teacher could "learn" you English using only English. ;)

Incidentally, we call teaching someone a foreign language by using nothing but that language "immersion". While it's certainly not the most gentle method in terms of initial learning curve, it has been demonstrated time and again to be among the best methods. It works on adults to. If you move to a country without knowing the language, the desire to stay alive (and eat!) tends to be a very strong motivator for accelerating the learning process.

(Compiling the source of a compiler with the compiler itself is a mark of maturity. Writing a specification in itself doesn't seem in any way outlandish.)

Re:The More they add, the less I like

[EugeneK]

Way to miss the parent's point! His point was exactly that; that if you use CSS you don't need to send the styling info with every page load.

Re:The More they add, the less I like

[porneL]

IE6/7 have all those weird box model problems with XHTML 1.0

No, IE can handle box model perfectly. It's XHTML it can't handle at all. You must be sending your pages as HTML (text/html) and you've put XML prolog in your HTML, which triggers quirks mode (you may think it's XHTML, but browsers see it as HTML with lots of syntax errors and bogus DOCTYPE).

obligatory hixie reference

Re:The More they add, the less I like

[cultrhetor]

Actually, you can just add a stylesheet that makes sure it renders correctly on PDAs. No need for switching:

<link rel="stylesheet" href="styles/standard.css" type="text/css" media="screen, projection" />
<link rel="stylesheet" href="styles/pda.css" type="text/css" media="handheld" />

Re:The More they add, the less I like

[Bogtha]

To be honest I didn't read the spec (yet) but it appears the page is following HTML 4.01 strict but the doctype is incorrectly specified.

Well read the spec then. That's the HTML 5 doctype. The only reason they use a doctype at all is because otherwise it would trigger quirks mode.

Re:The More they add, the less I like

[dhalgren]

It's what you did during Stairway to Heaven in grade 10.

Unless you're a Slashdot regular, of course.

Torben

Re:The More they add, the less I like

[eclectic4]

"this incessant pushing of the technology/standards envelope is creating a lot of disjoint, stilted, and otherwise unreadable web sites. etc..."

The technology standards didn't create those websites, the developers did. You seem to be asking to halt technological advancement to save developers from themselves, when it should be the other way around...

Re:The More they add, the less I like

[caudron]

There's no such thing as a "box model" in XHTML 1.0. The box model is a feature of CSS.

I agree with everything else you said, but have to defer to the other guy on this one.

CSS allows you to play with the box properties (like borders and padding and margins), but the box model is the direct result of the div structure of XHTML 1.*. I know why you say otherwise. Conventionally, when we talk about the box model, we are talking about CSS's use of it, but technically, convention is wrong, in that the box is defined in the XHTML rather than the CSS.

God. I am such a 'tard that I couldn't let this minor point go uncorrected. lol! Forgive my pedantry. ;-)

Tom Caudron
http://tom.digitalelite.com/

Re:The More they add, the less I like

[MP3Chuck]

If the user wants it to be up to themselves, they can disable external stylesheets and define their own. The whole point of semantic [X]HTML, with presentation broken out into CSS, is that the user can do that, if they so choose, and still have a functional document in front of them. The other 99.999% of Web users would likely perfer if they didn't have to do that.

Re:The More they add, the less I like

[BlueStraggler]

doing a decent layout with just HTML requires tables which in the end creates tag-soup because they weren't designed for page layout. CSS eliminates that problem.

Apparently you haven't actually tried to do real layout with CSS. There's only a small number of basic layouts that actually work, and even those may require ridiculous hacks that exploit browser bugs to hide rules from certain browsers, and use image backgrounds to create the appearance of cells. Google for 3-column page layouts to see the contortions you have to go through to make a basic 3-column layout with one liquid column. And that's a pretty damn trivial layout. Anything more complicated than that and you're in for a world of pain.

Small wonder then, that everyone seems to have standardized on 2- and 3-column presentation. It's pretty much the most you can do without a Ph.D. in web voodoo. But let's not pretend that this is a holy advance in the world of design. It's the opposite - a set of handcuffs for the designer, for the sake of getting one useful feature (namely, separation of content and design).

To be fair, I am referring to CSS as it has been implemented, not as it has been designed. There are a ton of fabulous CSS features that address all my complaints. It's just that none of them work out there in the real world.

And meanwhile in IE Land...

[CastrTroy]

And meanwhile in IE Land, we're still trying to get proper CSS Support. It will always come down to the lowest common denominator, especially when the LCD is the most popular browser. Nobody is going to code HTML 5 pages when the most popular browser doesn't support them. It's great that MS has finally made some headway with IE 7, but if they wait another 5 years until their next major release, then they are going to be even further behind. While all the other browsers are working on CSS3 and HTML5, MS is still working on CSS2 and HTML4.

Re:And meanwhile in IE Land...

[Excors]

A fundamental design principle of HTML5 is that HTML5 pages should work in existing browsers like IE6. You can write <input type="email">, and an HTML5 browser will allow useful auto-filling and immediate validation feedback, while old browsers will simply show a text box. New elements like <video> can have fallback content, e.g. to embed a Flash video player. New elements like <canvas> can be partially implemented by JavaScript in IE6. The HTML5 doctype (<!DOCTYPE HTML>) is chosen so that it triggers standards mode (as opposed to quirks mode) in all existing browsers.

So, you can write in HTML5 to provide added benefits for users of browsers that understand HTML5, while still being no worse than before for users of older browsers. And given that IE development has started up again, with IE7 making some progress on CSS standards compliance, and given that Microsoft is a member of the W3C's HTML working group which will almost certainly accept the current HTML5 work (as it was the reason for the working group to be formed, and nobody has raised any serious objections since it was proposed), I believe there is reason to be hopeful.

Re:How will this effect IE7

[aicrules]

No one is compatible with HTML5.

I started to

[caluml]

I started to code my pages in XHTML. But it's just not worth it. Use what works. :)

Re:I started to

[jd]

We need to replace the blink tag with a more general annoy tag that can have blink attributes. This will allow W3C to properly codify all the ways of being a complete pain under a single, unified system.

5???

[aicrules]

I just figured out HTML1 and I am still crying that doesn't work! :~(

Update CSS not XHTML/HTML

[oliverthered]

What we need is an updated version of CSS that lets you do things like reference other elements attributes so that you can create tables and line up things across/down the page. The ability to put different images on the left and right hand sides and top and bottom and all variants off would be great for putting rounded corners on things etc... instead of having to do hacks link putting in extra p tags just for the image.

HTML is more or less fine, give me a better version of CSS anyday.

Re:Update CSS not XHTML/HTML

[drinkypoo]

IMO what we really need in CSS is variables and math. Variables are really key. And to be able to say that the width of an element is n% of the width of another element, even when it is not nested within that elements, is also key - otherwise you have to use javascript for assloads of things. Of course other similar things would be possible. This is absolutely critical. Without it CSS makes life harder in a disturbing number of situations.

Re:Update CSS not XHTML/HTML

[richdun]

Some of that is in the CSS3 and further specs, like the advanced layout module, but those are beyond the reach of even the latest versions of FF, Opera, KHTML, etc. at the moment.

But, really, XHTML 1.1 is a great standard, and instead of moving ahead, let's try to get everyone to use it first. It hasn't been updated in forever (forever in web terms, of course) because the push has been to get everyone to actually use standards, and to get browser support of CSS2 and eventually CSS3 complete across all platforms and engines.

Just glancing over it, the HTML5 standards up at WHATWG worry me slightly. There seems to be a lot fo presentational/non-structural markup sneaking back in. Not necessarily as obvious as some of the older tags that were dropped in HTML4/XHTML1, but still. We have to keep in mind the separation of powers - XHTML/HTML for markup, CSS for presentation, and DOM for scripting - or things will just get way too complicated again.

Make things easier and more accessible for the developer/design? Sure. Add presentational content to HTML so he/she doesn't have to learn how to properly use CSS and the DOM? No. Do this, and it'll open the floodgates for everyone (MSFT) to add "special" tags to further "help" the developer/designers. Next think you know we'll be running around with a bunch of "Works best in ..." graphics like its 1998 again (only this time we'll be using PNGs or JPEG2000s instead of GIFs).

Re:Update CSS not XHTML/HTML

[TheRaven64]

I'd also like to see something like XSLT supported better. I hate having to put lots of class="foo" attributes in. Just let me define new tags and have the browser translate them into something sensible with a simple mapping. It would reduce bloat a huge amount.

Re:Update CSS not XHTML/HTML

[richdun]

Sorry for the double-reply - hit Submit before the thought completed in my head.

Advanced selector support in CSS3 could also help with reducing the need for tons of classes and such. But that's also an implementation issue - Safari is pretty good, Mozilla and Opera are getting better, but even IE7 doesn't support all the CSS2 selectors, let alone the CSS3 ones.

For those not familiar, I'd encourage checking out the full list of selectors - especially the nth-child and attribute ones, which could make a huge difference for paragraph spacing, drop caps, and, most commonly, the "zebra striping" in data tables, forums, email apps, etc. Check out Andy Clarke's Transcending CSS (http://www.transcendingcss.com/) for a really great look at advanced CSS3 features that will make you hate browser manufacturers even more and make you wish all these things were actually usable in today's web.

Re:Update CSS not XHTML/HTML

[WebCowboy]

What we need is an updated version of CSS

What we REALLY need is updated versions of BROWSERS (especially IE--even v7 is out of date in terms of compliance). We don't need CSS4 or HTML5 or XHTML2...browsers still have a hard time with CSS2, HTML4.x/XHTML1.x!

do things like reference other elements attributes so that you can create tables and line up things across/down the page.

Does not CSS3--a current standard with which nobody yet complies--provide means for columnar layouts? Also, last time I checked the latest HTML/XHTML specs supported table, tr, th, td etc...and none of it is deprecated...and even CSS2 has tabular layout stuff. It drives me up the wall when I see something like a calendar or other tabular information marked up with silly CSS an div and span tags all over the place. The thing is, if it is "semantically" a table, you should USE an (x)html table! Not only is the CSS simpler, the actual content (in terms of (x)html source) makes more sense to parsers and humans.

OTOH, is you want columnar layout for non-tabular data (news sites, discussion forums) then stop using tables already! It isn't all web developers' faults...the biggest problem is that the lack of support for CSS3 and the difficulty in doing such layouts in older versions of CSS is reinforcing bad behaviour (however, there is NO excuse for marking up lists with divs and spans--another annoyance to me. The ol, ul and li tags are there and CSS supports formatting these elements quite well).

HTML is more or less fine, give me a better version of CSS anyday.

I mostly agree if you mean strict XHTML, though CSS3 is a big improvement as well and what we need is better CSS SUPPORT. A few things I'd more like to see to make the WWW a better place:

* better native support for SVG in browsers...if you want rounded corners and other spiffy presentational elements THAT is probably the most appropriate place for it to be done.

* if (x)html is to be messed with lets look at how forms are handled...we could do with something that is more "evolutionary" than what W3C has been mulling perhaps.

* the standard that needs the most work is Javascript/ECMAscript...the xmlhttprequest object is not implemented as a standard and there is no standard proposed for it. Instead we have a proposed standard for similar functionality by the W3C (DOM level 3 "load and save") that is more cumbersome and nothing like the "de-facto standard". This "real" standard (DOM3LS) thus far is only supported in Opera IIRC, and the Mozilla team really pushes back when asked to work on such support. People, pick something and get it working the same in all browsers..this is what holds back "web 2.0" the most IMO!

Re:Update CSS not XHTML/HTML

[FooBarWidget]

The only reason why I don't actively use XHTML (and instead use HTML 4 strict) is because Internet Explorer doesn't correctly support it. It renders XHTML as HTML 4 *transitional*, which activates IE's broken 'compatibility' box model. And IE doesn't support the XHTML MIME type.

The past new years I've noticed that the browser world has become IE versus "the rest". Firefox, Opera and Konqueror seem to render everything nearly identically. 95% of the rendering problems during website development seem to come from IE, and IE only.

This could be the leverage needed against MS

[erroneus]

The biggest problem/complaint against Microsoft is that their dominance is hurting standards. Perhaps to some degree, the standards body could come up with a way to force Microsoft into being compliant and compatible? Perhaps there should be a level of completeness of implementation that would be required before being approved as "HTML5" compliant or compatible?

We know Microsoft is capable but they just don't want to. Their weight and sloth is an abuse to the community at large.

Re:This could be the leverage needed against MS

[starwed]

I'm sure it will ease your mind to know that the chair of the working group works for MS... :P

Re:This could be the leverage needed against MS

[AthenianGadfly]

If I were a chair, I would not want to work at Microsoft...

It could be dangerous

Opera

[RonnyJ]

Speaking of Opera, version 9.2 was released yesterday, but doesn't seem to have warranted a headline here as of yet.

http://www.opera.com/download/

Talk about spin!

[Bogtha]

"The World Wide Web Consortium (W3C) has been slumbering the past several years

No, the W3C have been very busy.

XHTML was last updated in 2002

No, XHTML was last updated two months ago.

no one is taking seriously their largely incompatible work on 'next-generation' XHTML or 'modularized' XHTML.

Everybody is ignoring XHTML 2.0 because it isn't finished yet. XHTML 1.1 is not an option for most developers for one reason in particular: you can't use it with Internet Explorer. Blame Microsoft.

Both HTML and XHTML are in sorry need of removing deprecated items

No, both HTML 4.01 Strict and XHTML 1.0 Strict are available for those people who wish to use a document type that doesn't include the deprecated stuff. And even if they weren't available, nobody needs deprecated items to be removed. If you don't want them, don't use them. Just because they appear in a specification it doesn't mean you are forced to use them.

The quality of this work has reached the point that Apple, Opera, and Mozilla have requested the adoption of HTML5 as the new 'W3C Recommendation' for Web development.

No, they are requesting that the W3C — the organisation you've just written off as closed and useless — adopt their work as a starting point, so that it can be developed further at the W3C. They aren't asking that the W3C give it Recommendation status, they are asking the W3C to take over its development.

Re:Talk about spin!

[Anonymous Coward]

Everybody is ignoring XHTML 2.0 because it isn't finished yet.

I thought it was because it was a pointless and unneeded reformulation of existing standards with no BC?

XHTML 1.1 is not an option for most developers for one reason in particular: you can't use it with Internet Explorer. Blame Microsoft.

1.1 is not an option if you want to support UA's that only accept text/html and Lynx will never support application/xhtml+xml. All XHTML1.1 does is modularize version 1.0, most users probably don't even know what that means ;-o

Re:Talk about spin!

[Bogtha]

I thought it was because it was a pointless and unneeded reformulation of existing standards with no BC?

You're welcome to that opinion, but I think the fact that it's a work-in-progress is the relevant factor to consider when wondering why people aren't using it. Even the W3C themselves don't want anybody to use it yet. In their own words, from the top of the latest specification: "It should in no way be considered stable, and should not be normatively referenced for any purposes whatsoever."

Lynx will never support application/xhtml+xml

Lynx already supports application/xhtml+xml. According to the changelog, support was added almost three years ago.

Re:Talk about spin!

[Excors]

I think the fact that it's a work-in-progress is the relevant factor to consider when wondering why people aren't using it.

That's not a relevant factor for the Safari developers to say "the HTML standards process has been moribund; the W3C's HTML Working Group has focused almost exclusively on XHTML2, a new standard that was highly incompatible with existing practice" and "We declined to participate in the XHTML2 Working Group because we think XHTML2 is not an appropriate technology for the web". As far as I am aware, Mozilla, Opera and Microsoft are all not planning to ever implement XHTML2, whereas they are already working on HTML5 – HTML5 also has many features that are work-in-progress and which nobody is using yet, but which the browser vendors are already implementing, because they are valuable changes and don't break compatibility with the current hundred billion documents on the web.

Misses the point

[starwed]

The article misses a pretty large point: the w3 has already decided to work on the next version of HTML. The post linked to is a recommendation that the HTML 5 spec be used as a starting point for that work.

Re:How will this effect IE7

[Anonymous Coward]

Chris Wilson of Microsoft's IE team is one of the co-chairs of the HTML WG. The work of the WHATWG however has been created without any input from MS. It remains to be seen whether or not MS will adopt everything. There are also internal quarrels among the other three, read the public-html list archives to find out.

A bit premature?

[DragonWriter]

Maybe its just me, but I think its a good sign that a proposed spec isn't ready for adoption when it contains this warning on one of its elements (see 5.4.1 The UndoManager interface):

This API sucks. Seriously. It's a terrible API. Really bad. I hate it.

Its also not a good sign when it has sections with a note of the form "Does anyone know enough about $foo to write this section" or "Need to write this section". Certainly I can see a need and utility for something like the Web Applications 1.0/"HTML5" standard, but it certainly doesn't seem ready for adoption as a Recommendation yet.

Re:A bit premature?

[ZachPruckowski]

Yeah, slashdot summaries can be a bit misleading at times.

Why Developers Aren't Caring Too Much

[moore.dustin]

We are all going to be making a majority of our sites to work in IE6 for many years to come. The release of IE7 did hardly anything to change how I design my pages. All it did was add another browser to test in really. IE6 will remain on old Windows OS's (2000 cant run IE7) and non-upgraded machines; therefore, we will all develop for them as we have for some time now.

All this means to me as a developer is that I have another thing to keep track of in regards to my industry. Add it to the list which includes: Seeing if AJAX, RoR, and other Web 2.0 fads survive the next year; if PHP has even a glimmer of hope; PNG issues; content delivery to mobile devices; and of course assorted security issues.

We always have something new coming down the pipe, but that does not mean it is the next new thing. Many sing the praises of AJAX, but really it is far from perfect and likely to be replaced by something much better very soon. Nature of the biz I suppose, but I would not have it any other way !

Re:The Good News

[JordanL]

Moving away from pages that "only work in IE" would take a lot more than HTML5. Microsoft made IE the engine for displaying all system windows in their OS... when you browse my computer, what more or less happens is Windows generates an HTML DOM object to display the contents.

This is the reason Microsoft has had so much trouble with standards and the reason they will never be able to fully support standards as long as IE is integrated in Windows to the level it is. Standardizing would leave the OS itself high-and-dry, which is something Microsoft is not willing to do.

This was part of the reason Netscape sued them, and it was part of their anti-trust suit here in the US. Everyone knows this was done almost for the sole purpose of using MS controlled platforms (IE) to prevent non-MS controlled platforms (the web) from abstracting out all the functions a user needs. Most people still think the internet doesn't work on Linux, or that Linux has a "different internet" of its own. I would venture to say that their anti-competative practices with IE are the ONLY reason they still command the market share they do.

Talking to Microsoft before formalizing a web standard is like going to them for an open document format: you just end up knee deep in shit that Microsoft doesn't genuinely believe in the first place.

Horrible

[Zenethian]

This is horrible. The mess of backwards compatibility on the web, particularly in HTML, is what causes so much "liberal" output from web developers and designers. XHTML took a solid step forward in squashing some of those problems by creating a very rigid set of rules to be followed for document markup. XHTML2 addresses the actual semantics of it. Backwards-compatibility is not always a great thing. Something like XHTML2 promises a clean breakaway from the horrors of HTML. This "HTML5" seeks to make the web even worse off than it already is by providing developers and designers a free ticket to make their code as horrifically nonstandard as possible. Documents should consist of well-formed markup that is easily parsed by both humans and machine alike. From a purist point of view, XHTML is blissful, though it does truly have it's own set of issues. From a realist point of view, there will sadly always be backwards compatibility on the Web, but can we *please* restrict it to software implementation and NOT in standards?! The web browser will always need to support HTML 2, 3.2, 4.0, etc etc (not that they do, but they should...). If someone wishes to code in HTML 3.2 then they should. But the next version of the (X)HTML standard should not promote backward-compatibility. It should move the technology forward instead of accommodating for previous bad practices.

Re:microsoft?

[msh104]

"Fact is if MSFT doesn't make the "standard" MSFT won't support it properly."

that's exactly why they should be in the standard creation team.

Please, give us better layout tools

[PapayaSF]

Tim Berners-Lee, bless him, didn't seem to understand that anyone would ever want a web page with more than one column. So some genius (a name I've forgotten) thought of using tables for layout, and many problems were solved: multi-column layouts with headers and footers which stretched to accomodate content and rendered the same way (more or less) across all browsers and platforms. Hooray!

Then came CSS: coding could be much cleaner and more flexible, but tables-for-layout was considered bad, and we began wrestling with creating layouts using divs and clears and floats, having to use such kludges as negative margins in order to replicate table-like behavior. It can be done, but it's harder. So for HTML5, how about setting aside creating new but not-very-helpful features like "overline" (who uses that?) and coming up with things that actually help us create web pages? Why not create a tag called "grid" that acts like a table, but is designed for page layout? Most graphic designers use grids, and it would really help web design as a whole if something like that existed for us.

How about a way of having content reflow from one column to another when a window is resized? Page layout programs have done this for 20+ years, so shouldn't it be possible for a web page and a browser today?

So please, HTML5 people, don't just talk to computer scientists and advocates for the disabled when creating this new specification. Think of the people who actually have to lay out pages!

Re:Please, give us better layout tools

[apathy maybe]

If you have data that uses a table, use the "table" tag. If you don't, use CSS. HTML is not for describing presentation, that is what CSS is for. As such, your idea for a "grid" tag, is not really for HTML at all.

What happens when your page gets displayed on a phone? With CSS you can simply revert to a single column (or the phone can just drop the CSS), with "grid", you need two pages, one for desktops, and one for phones.

I think XHTML is fine, it works and does the job. The only thing I would like is a client side include. Apart from that, I think CSS needs updating, not (X)HTML (or perhaps just browser support for CSS?).

Re:Please, give us better layout tools

[Kelson]

Why not create a tag called "grid" that acts like a table, but is designed for page layout? Most graphic designers use grids, and it would really help web design as a whole if something like that existed for us.

Because this would make sense, and help web designers/developers actually do what they've been doing all along. Its just the people in the business of writing the spec, have something different envisioned.

Not really. If it is functionally equivalent to TABLE, then it's redundant markup (like the old MENU and DIR list types, which were in practice equivalent to UL). It'll also have exactly the same shortcomings that table-based layouts have (particularly: mixing presentation in your structure, and limits on scalability, particularly going down to small devices like phones). The only thing that will distinguish it from TABLE is that parsers will know not to interpret it as tabular data. You may as well add a "NOTDATA" attribute to TABLE.

The only designers it will benefit are those who follow the "don't use tables" mantra as received wisdom, rather than understanding the reasons behind it. It's just like people who try to use CSS to imitate a table layout in order to present actual tables, because they've heard "tables are bad, use CSS instead" instead of "tables as layout lead to a number of problems with can be avoided by using CSS instead"

Re:Please, give us better layout tools

[tuxedokamen]

While I agree that page layout on the Web is no where near as simple as it should be (I'm doing a webapp right now, and sometimes it really does feel like alchemy), I disagree slightly with your last point. It's vital to listen to advocates for the disabled--particularly the visually impaired and those with motor difficulties--when solidifying standards like this. The whole idea of the web is accessing information, and if a designer takes the easy way out in terms of implementation and leaves huge swaths of people unable to see the content, IMHO they've kind of missed the point entirely. This is, in part, why tables, spacer gifs, etc. are bad, as I understand it (and also, from my own testing): they really play havoc with screen readers, to the point that a page is nearly incomprehensible. And that assumes the screen reader can make enough sense of things to read them at all--if a a page is malformed enough, there's just silence. I'd imagine the situation with braile-output devices and systems for navigating by alternative means are similar.

Government websites have to be fully accessible according to a government standard (508, I think), but non-government websites certainly don't. And making the argument that the visually impared/those with motor control difficulties make up such a small percentage of the population is not only superfluous, as even ten percent of the total internet population is well over 100,000, and the number of people with alternative access needs will only continue to rise as access proliferates, but outright discriminatory. If the guilding principle of the web is information and communication for all, exclusion isn't an option.
(I realize I'm not using exact statistics here, but I unfortunately don't have time to look them up right now.)

For the webapp we're doing, our team is adhering to the Authoring Tool Accessibility Guidelines (ATAG) 1.0 (http://www.w3.org/TR/WAI-AUTOOLS/). This set of guidelines and the Web Content Accessibility Guidelines (WCAG) are perhaps more important than entanglements over HTML5 and such right now, as they're concerned not only with implementation (how a page is programmed), but how it is designed, from the ground up, to be accessable to all. Implementation should flow from this, IMHO.

Re:Firefox plugin

[TheRaven64]

There is a Gecko ActiveX control. It is, theoretically, possible to detect IE and send a page that includes the ActiveX control and runs the rest of the site in that. It's a several MB download though, and people might get bored waiting for the site to load (not to mention your bandwidth bills).

Re:I'd settle for some taking away

[Bogtha]

I agree that it's silly that they are required attributes, but merely missing the attributes off doesn't dump you into quirks mode. Quirks mode is determined by the doctype you use.

Multi-column is already in the pipeline

[Kelson]

How about a way of having content reflow from one column to another when a window is resized? Page layout programs have done this for 20+ years, so shouldn't it be possible for a web page and a browser today?

The CSS3 multi-column module was designed for exactly that purpose. It's available in experimental form in current Mozilla-based browsers (Firefox, Seamonkey, Camino, etc.), and according to that page, it's available in nightly builds of Webkit, which will eventually become a future version of Safari. (Since the spec isn't final, the rules use -moz and -webkit prefixes, so that if the spec changes they won't have to change the official rule's behavior.) No word from Opera, though there are reportedly a bunch of CSS3 features in store for the next major update, and of course, who knows how long before we'll see it in IE.

Remember: HTML for structure, CSS for layout.

Re:Multi-column is already in the pipeline

[HTH NE1]

The thing about multi-column layout is it is very annoying to read when the columns are taller than the browser window. You reach the end of one column and you need to scroll back to the top of the next. It's the same annoyance as reading lines of text which are wider than you screen, just less frequently. (It works for newspapers and magazines because their whole page is visible and is only a matter of scanning with the eye.)

The proper rendering of multi-column text is to embrace horizontal scrolling and forbid vertical scrolling. Column width and gutter need to be an even divisor of browser width and leave column height and count to flow to fit the window.

Of course, this is at right-angle odds with the design philosophy of web pages, so multi-column sections pretty much have to be subsections of the page itself as like an IFRAME, allowing the multi-column view without disrupting the rendering of the rest of the page.

As a result, it's a design mess and people go back to the single column view with sidebars for additional information, using web design to its strengths rather than forcing it to behave like other media.

If it must exist, it should be a style sheet rule-set so it can be applied according to output type such as hard copy where that presentation makes sense. I'd prefer to have image masks that enable text to flow around the curves of an image, and leveraging transparency for it would reduce the bandwidth impact.

Explaining the plan

[alienmole]

Here's what I was thinking: ordinary users don't seem to have a problem installing Flash, which is a several MB download, when they're told that they need it to view a site. So if the Gecko ActiveX control does the trick, those of us who are serious about eliminating IE should detect IE visitors and display a page saying that you need to download the Firefox/Gecko control to use the site (or Firefox itself, of course).

Pretty soon, about as many people who have Flash will also have Firefox running inside IE, and it'll no longer be necessary for many people to cater to IE.

Re:I'd settle for some taking away

[J0nne]

you mean the following doesn't work?

textarea {
width:200px;
height: 100px;
}

I guess I've been doing css all wrong for years now :(...

Re:I'd settle for some taking away

[MankyD]

It works, but it's not standards-compliant - it's not required to work.

Re:Not to put words in his mouth...

[Bogtha]

Hover rules aren't useless eye candy. Hover rules are visual feedback letting you know you are over something clickable. If you move your cursor across a bunch of links, it's immediately obvious which one you are currently over without having to pay attention to precisely where your cursor is. Usability++.

Forget HTML, it's CSS that's Broken, deal with it

[StreetFire.net]

I WANT A REAL LAYOUT LANGUAGE!!!!!!!!

I've tried, I really have, to embrace the Zen garden Juu-Juu of CSS, can you make a simple blog page work in CSS? sure! Can you make an massive website with many different templates and variable width data-areas work in CSS? Yea, if you're a complete lunatic. but you have to get there with hack over hack over hack over hack. Here is the deep dark secret of CSS, it's not designed for layout. It's fantastic for styling, but try doing a Box-model or Float layout and you quickly realizing you're asking CSS to do things it wasn't intended to do, and it simply does not break gracefully the way a simple table layout does (You know floats were originally intended for pictures, not layout areas). So while I respect the purity of a CSS for style, HTML for content concept, in practice CSS is just as much of a kludge as Table design. I've saved hours of time and reached wider audiences of compatibility by going for a hybrid design, but this breaks the "standards".

IMO, standards should follow simple elegant solutions, a hundred lines of CSS browser compatibility code and float hacks is far from an elegant solution. PLEASE PLEASE PLEASE give designers a proper layout language!!

Dear W3C,

[circletimessquare]

Please revive the <BLINK> tag. I thought it was as awesome as MC Hammer. In fact, just go ahead and make an <MCHAMMER> tag.

What would it do? You have to ask!?

Yours,
the early 1990s

Today is NOT a good day to die.

[ScentCone]

When was the last time you even saw a computer that had IE5 on it?

So, I've got a client that runs an e-commerce site. At least a couple hundred orders per day. I did a quick dig into today's stats. So far: 4 orders from people running IE5, and one from a Netscape 4 flavor. All appeared to be on dial-up connections. A little over $1600 worth of business in those 5 orders. These are orders for non-essential items, which suggests disposable income that COULD go into a computer upgrades, broad-band connections, etc. for those shoppers, but which have not. I absolutely guarantee that my client would rather have today's business from those 5 customers than have whatever liberty may come from being able to leverage current formatting fanciness/compatibility. Their site renders just fine in every browser to date, and that $1600 is in the bank, instead of that of a CSS-ed-to-the-hilt, hipper-than-thou competitor. Someday the numbers of legacy users will drop low enough to warrant the change, but $1600 before lunchtime says today's not the day.

Re:Today is NOT a good day to die.

[lahvak]

He spent $6k once to make $1.6k a day.

Re:Today is NOT a good day to die.

[drix]

Meh. First, you're talking gross. What's his margin on those $1600 worth of orders? If it's "standard" retail, let's say 5%. So $80 bucks a day. Now, how much extra time and effort did it take you, the developer, to support browsers that are almost a decade old and that, by your own admission, affect roughly 2% of the userbase? My guess is at least a couple thousand dollars, unless you adopted a lowest-common-denominator approach, in which case the site must look unappealingly 1997. More importantly, what sort of tradeoffs were you forced to make? Have you studied at all how much business your client is forgoing by not leveraging the current "formatting fanciness"? Here are a couple points to consider.

1. People like sites that are clean looking and easy to use. Marketing studies have consistently shown that people will pay more for the exact same item from a place that sells it in a more aesthetically pleasing manner. (I'm not saying this can't be done with HTML 3.2 or whatever, but it's much harder.)
2. Standards-compliant sites that use semantic markup place higher in search engines, netting more impressions and more sales.
3. Table-based layouts are slow and unresponsive. How many people here remember good old NS 4 sitting there blank-faced, cranking away on the old, complicated table layouts of yore. I do. Responsiveness is huge; people have come to expect it as the rule, not the exception--a marked departure from the dark days of IE5 and NS4.

Re:Today is NOT a good day to die.

[ScentCone]

First, you're talking gross. What's his margin on those $1600 worth of orders?

He makes about 35% markup on orders like that. As of a more recent check (some hours since I chimed in on the thread), the ancient-browser-checkouts have now grossed about $2200. We'll call that, conservatively, about $500 of profit (not counting taxes, flushing the toilets, pizza for the warehouse guys, etc).

the site must look unappealingly 1997

No, I'd say it looks more early-2005. The design is deliberately lean, spartan, and surprisingly navigable considering they have around 12,000 items. The are leveraging Froogle, affiliate marketing with feeds and hot links into products... lots of the more recent goodies. There are some nested tables in play, still... but they come back with first-page Google hits on a great deal of what they talk about and sell.

Responsiveness is huge

Yes, it is. But any latency I've had to fight was almost always due to database performance problems, usually because some session management table or other beastie had outgrown the way the indexes were built, etc. Believe me... a complete redesign for new standards is desireable, and could indeed bring in some otherwise missed sales. But it's nice to not run off the little old ladies and their credit cards, too. 10 of them today, it looks like. That's about 300 of them per month, and they do a lot of repeat business... the business has about a 45% repeat customer rate. Which might not sound great until you realize they're growing rapidly. So, don't "meh" something that's working pretty solidly, and which is very much a topic of discussion and planning at the business. My point (back to the thread, here) is that the "web designer" to said "when was the last time you even say a machine running IE5" is full of crap. I'm not just seeing them, I'm seeing them show up and spend money.

Re:Today is NOT a good day to die.

[drix]

Seeing as we're a mouse click away from forming our own judgements.. what's the site?

Re:Today is NOT a good day to die.

[businessnerd]

Probably better to not post the link. From what has been described, I doubt the site could withstand a slashdotting.

Re:How will this effect IE7

[maxwell demon]

OTOH, any bets on whether IE7 would support it ... ever?

No. IE7 is already out now, and MS would be silly to give the next version the same version number as the current version.

Think of our bottom line!

[Wayofthebit]

I get the feeling that the W3C and these three businesses don't really give a damn about the web or its users. All they care about is the bottom line or justifying their very existence.

From reading their faq and some other articles. It seems they are saying they don't believe in using XML on the web. They are willing to support the cludge use of HTML and XHTML in a more standard way. However, they never intend to become complete XML solutions. I thought this was why XHTML was created in the first place. To begin the transition to complete XML solutions on the web. If they never intended to go the distance why have they waited so long to tell their users?

In the end it just sounds like they want to keep things the way they are now. Just try to agree on a better way of supporting the various markup we have now. They've had years to get HTML to be consistent. Yet, they've not been able to acheive this seemingly simple task. Why would I believe HTML 5 to be any different? XML is supposed to be more precise and standardized. Perhaps that isn't the case, but they sure as hell waited long enough to tell thet rest of us.

Why HTML5 rocks?

[porneL]

<!DOCTYPE html>
<meta charset=utf-8>
<h1>Hello world!</h1>
<p>This is a complete, valid HTML5 document

and it does work in current browsers as intended, even in IE6.

WRONG!!

[mabhatter654]

Any serious team should kick IBM, Adobe, and Microsoft out. Period. Those companies are the ones that wrecked the current W3C specs by bloating them up with "good ideas" from their established products, then not doing good implementations. MS was in meetings for CSS2 and XHTML, way back in 1999 and they have yet to properly support it. The mother of all W3C specs is SVG... the committee bit when Macromedia and Adobe wanted everything + kitchen sink, but the spec is so ambiguous and bloated nobody can implement it.. worse browsers don't implement the SAME features so it's even more pointless.

I'd like to see JUST browser makers and web designers in on the next specs. I under stand TBL (father of the web) doesn't like the idea of "web apps" over semantic documents, but the case is lost. The biggest thing is that the W3C doesn't actually make a fully useful browser of their own... they should defer to those that DO make browsers and those who design web pages and create a spec that's 100% useful and implemented rather than "pie in the sky".

CSS is a failure, as is much of "Web Design"

[Ralph Spoilsport]

Adobe cooked up PostScript. Theoretically, you could "read" PS code. It was hard, but not impossible. Within a few years this PAGE DESCRIPTION LANGUAGE was made simple and invisible by programs such as Illustrator, FreeHand, Fontographer, then Pagemaker and Quark as well as other apps that are gone and largely forgotten (ReadySetGo anyone?).

You place an element on a page. size it, etc. and you have NO contact with the code underneath, and frankly, you DON'T want to deal with it. It's messy and complicated.

Now, HTML was cooked up YEARS ago, and the closest we have to Pagemaker/Quark is Dreamweaver. Brilliant. And placing elements on a page is such a hack job between browsers that an entire industry of useless coders have sprung up to "take care of that" for us - CSS specialists who demand serious dollars to do something that shouldn't even be handled by humans. And WHY is this so?

1. Microsoft Their insistance on being non-public standard compliant and shoe-horning more crap into their browser and DOM for IE(x) has been a monstrous thorn in the side of web developers everywhere. At the same time, it is this incompatibility that gives CSS specialists and web designers/developers a job. Still, this is not a happy situation, and it is not going in a useful direction (HTML 5 will be gleefully ignored by Redmond, and the complexity of the resulting situation will only give more work to more CSS/AJAX specialists.

2. Fiefdoms The resulting incompatibilities that create the need for specialists actively works against finding automated solutions. If web design was reduced to the level of Adobe InDesign and web development was made as simple as visualBasic, then much of the "Web Industry" would disappear. Thousands would lose high paying jobs. I nreality, they (and this means YOU) are nothing but parasites subsisting on a faulty broken technology. Fix the technology and the cost of development would gradually collapse. Insisting that a web designer should know code would be seen as absurd as insisting that a type designer know how to read PostScript or raw TrueType instructions.

3. Flash et al. It seems much of the web started as one thing and ended up another.

Tables were built for tabular data, they soon became the structure by which a page was architected, as it put things in a specific place in page that was pretty much the same across browsers. And that is still the case. If I'm going "whip out" a page with a variety of elements arranged in the page, I'll use a table. It's faster and easier and it works. It's not as flexible or useful as CSS, but it works. With CSS, you spend huge amounts of time tweaking crap to appear across browsers. Still, Tables are old and not as flexible, and have been pressed into doing service they weren't designed for.

Same with CSS. It was developed to make things all purty-like. Now it's used for page layout and element placement - WAY beyond it's original mandate, which was taken from Desktop Publishing (stylesheets).

Flash was just a way to do spline based illustrations and animations on the web. When it was clear Flash could do much of the basics of what Director could do in a tiny fraction of the footprint, they began "directorfying" Flash, and shoehorned a Lingo into it, known as ActionScript. It is now a full blown frankensteinian disaster that is considered a "development platform", which is a cruel and misdirected hoax.

The examples are many and continuous. The "web" is a hack job. It is a slow moving train that has left the tracks (IE5 did that), and is slowly dying in its own complexity (of CSS, XHTML, AJAX, etc. hackery) as it rolls in excruciating slow motion off the cliff.

And now that the very livelihoods, mortgages, and private schools for the kids are DEPENDING on this complexity, the only logical conclusion is increased complexity, resulting in an ever higher range of exclusivity, casuistry, and sophistry. It will die, replaced by a lower context system, and when it does, a lot of people are going to get hurt. And, given the nature of the audience on /., it will mostly fall on their shoulders.

RS

Re:Forget HTML, it's CSS that's Broken, deal with

[GigsVT]

You are very correct. CSS gets much more hacky than "legacy" layout if you try to do any significant layout with it.

I tried to make a simple 3 column table with CSS only. After struggling with that for an hour, I said fuck it, and put an old style table in there. It was much easier.

Okay, but....

[Kadin2048]

Okay, but just to play the devil's advocate here, why should that be part of the page design, and not a feature of the user's reader, perhaps as a configurable option?

It used to be pretty standard for people to customize their browsers in order to change the text, link, followed-link, background, hightlight, and other colors; why does the page designer necessarily know better than the users themselves what the user wants?

We've moved a long way over the past few years towards making the browser into a generic 'portal' that simply displays whatever the web developer wants to toss up on it for the user to look at; frankly it's very television-like.

However, there is a completely different conception of the internet where the pages should be marked up as generally as possible, and the user's browser should then choose how to display the information in a way that's meaningful to the user. It would probably mean that "your Internet" wouldn't look anything like "my Internet," but there's no inherent reason why that's bad. We've grown to treat it as if it is, but that's only because we want the web experience to be like flipping channels on a TV, where your Discovery Channel looks exactly like mine.

Re:Okay, but....

[Bogtha]

why should that be part of the page design, and not a feature of the user's reader

With CSS, it can be both. The "C" in "CSS" stands for "Cascading". The style rules suggested by a web designer cascade together with style rules preferred by the user.

As for why it's a good idea for web designers to have this feature, well it's for the same reason any styling is useful for the web designer to have. Because although the user should have the final say (which CSS allows), it's difficult to predict exactly which presentation is most suitable for all the pages you are likely to encounter in the future. The web designer that produced the page, on the other hand, knows the context in which the information is being related and has a good chance of being able to come up with a more appropriate presentation than something that is generic to all pages.

However, there is a completely different conception of the internet where the pages should be marked up as generally as possible, and the user's browser should then choose how to display the information in a way that's meaningful to the user.

Well if that's what you want, then pages being "marked up as generally as possible" is exactly the opposite of what you want. If you are relying on the browser to handle all presentation, then what you need is specific, accurate markup, not general stuff.

In any case, the two goals are not mutually exclusive, and CSS has been handling this for over a decade. For those users who like to be in control, they can configure their browsers to ignore author-supplied stylesheets. Everybody else can take what is suggested by the web designer, or configure their browser to make only minimal changes (e.g. font size).

Re:Firefox plugin

[illegalcortex]

Note to self: When making next joke about a phony plugin whose purpose is to eliminate rendering differences between IE and Firefox having rendering differences of it's own, google made-up plugin names first. Also, be less subtle. ;)

Re:HTML5 === steaming pile of shit

[wootest]

Are you seriously suggesting that XML's error handling is ideal - that the correct way, after 17 years of accepting non-perfect content in web browsers is to inform readers of a page that a href attribute that contains an & that's not part of an entity like & that the page is broken and then quit without showing any information? XML is awesome for programs reading data produced by other programs. Draconian error-handling is appropriate in such scenarios. However, way too many people are hand-writing HTML - and it's way too hard to generate sensible HTML - to be able to make that plausible.

HTML5 defines and codifies separate HTML parsing rules, mostly backwards-compatible. You're able to download a HTML5 package and parse almost any site on the web today - are you willing to bet you can do that with half the XHTML pages? (Even if you still want to use XML, you can just write XHTML5, which is HTML5 with XML parsing.)

WHATWG's two specs specify both the HTML markup and the DOM. The two go hand in hand - that's why it in comparison to HTML 4.01 or even XHTML 2 looks like a "mish-mash of markup and javascript", because it's just not solely a markup spec anymore. If there's going to be a new HTML spec, it deserves to be a solid and holistic spec. The least you can do is include the DOM model inline, and not just publish a different spec, by a different working group, and brand it as "DOM HTML Extensions", or what have you.

Re:Not to put words in his mouth...

[Bogtha]

Links are immediately obvious in regular HTML pages because they're a different color and usually underlined. Your mouse cursor also changes when it is pointing at them.

Did you even read my comment? I'm talking about when you are moving through a list of links, like a navbar. Just because the links look like links and your mouse pointer indicates that you are over a link, it doesn't mean you are getting strong visual feedback about exactly which link you are over.

If you have a bunch of links back to back in such a way that it is difficult to tell where one stops and another ends, then you probably need to redesign your webpage anyway.

Take a look at your own page. You provide an unstyled list of links at the top. When I put my mouse over the first one and move it down over the list, it's not obvious exactly which one I am over when I am anywhere near the edges. I have to look at the status bar if I want to be sure.

This is not about funky layouts or designs that scew things up. This is about a totally normal situation — adjacent links — being slightly improved by the appropriate use of visual feedback.

Re:Forget HTML, it's CSS that's Broken, deal with

[Phroggy]

Thank you.

I don't really care what they do with HTML. As long as support for the old versions doesn't go away (and as long as you include the appropriate DOCTYPE there's no reason why it should), I'll always have the option of using the old version if I like it better, or using the new version if they make real improvements. I use the W3C validator, and a Firefox plugin to do the same (although the Mac version seems to be broken at the moment). HTML is great. I'm sure they'll make it better. Fine.

But CSS is a nightmare. I've got two books on CSS on my bookshelf; neither seems in any way comprehensive, and I don't think it's the fault of the book authors. My horribly broken stylesheets always pass validation anyway, because the syntax is fine, they just don't work the way I wanted. I'm not looking forward to building a new site a client wants me to make, partly because I know I'll have to build a new stylesheet for it. I love what CSS is capable of doing, I love the concept of using a stylesheet instead of splattering layout and style markup all over the HTML. But CSS, it its current form, is painful.

I want a AJAX tag :p

[VGfort]

Do some web 3.0 stuff for me

<curvedcorners>Bling Bling</curvedcorners>

and maybe they will implement the </sarcasm> tag, no opener needed :p

Re:Forget HTML, it's CSS that's Broken, deal with

[angryrobot]

Here's one method:
http://alistapart.com/articles/holygrail

Here's another:
http://www.glish.com/css/7.asp

But yeah, it's hard. Lot's of people have found solutions to the problems, though. Even if it is cumbersome, I think it's better than visual markup.

Re:I think this is great

[argent]

That be the general definition of CSS (for better or worse).

CSS separates style from content (that's right there in the name, cascading style sheets), it doesn't address layout at all, which is why people using it for layout have to come up with horrible hacks with floats and the like. They're no better than tables, and I'm glad -moz-column-width is ugly and prefixed and not a standard, because it's too damn specialised. Reminds me of ... too bad that wasn't <-mosaic-image ...> so there was a reason to switch to <fig>...</fig>

Validation...

[cyclomedia]

And therin lies the problem, the w3c make a big shouty shiney deal out of "validating" HTML and CSS markup in pages - wether or not it actually produces desirable output in the shitty attempts at support by ALL the browser vendors - and not copyrighting the terms "HTML" and "CSS"* and then not allowing browsers to claim to render them unless the BROWSERS ITSELF CONFORM TO THE SPECS

*yeah i know it's too late for that now but "HTML 5" could be called "WML : Web Markup Language" instead, whilst being new and buzzword-tastic, M$ could not then release a "WML Support Upgrade" for IE7 unless the w3c said so.

too bloody simple, obviously

Re:I'd settle for some taking away

[Bogtha]

I'd be happy if we just had a way to layout the page such that it would size with the browser window.

Er, we do, and it's actually the default. Load a page without any styling information at all and see for yourself. If you want to specify a width, use { width: 75%; } or whatever relative width you like instead of using px, pt or whatever you are using now.