HTML5 and CSS3 - Is there a Web standards divergence?

[kgun]

At present I am reading the Sitepoint book - HTML5 & CSS3 for the real world:

http://www.sitepoint.com/books/htmlcss1/

In the section

"Would the real HTML5 spec please stand up"?

There I read that there is now an additional organization "the Web Hypertext Application Techonology Working Group" WHATWG that serves the living HTML standard, dropping the "5" found at http://www.whatwg.org/specs/web-apps...ork/multipage/ . It is called a "living standard" because it will be in constant development and will no longer be refered to using incremental updates.

The W3C's version of the standard can be found at :

http://dev.w3.org/html5/spec/Overview.html

So what is the difference between the W3C spec and that of WHATWG? Briefly the WHATWG version is a little more informal and experimental (and, some might argue more forward-thinking).

WHATWG was formed by a group of people from Apple, Mozilla and Opera.

Should you be worried that there are two versions of the spec?

In short, no.

What is your opinion?

[jhannawin]

Having worked in the computer industry for some 20+ years, I have seen many standards divergence problems. Usually, a group with a restricted agenda propose a derivative of the core standard and pushes it to their particular interest group. Whether this derivative succeeds depends upon the strength and quality of the original standard, and the influence of group. Big names in the group do not guarantee success in the derivative, just look at the history of Unix to illustrate that one.

In this case the W3C appear to be doing a pretty good job. Maybe things don't move fast enough, but they have come through all of the crazy 'extensions' that Microsoft and others tried to force on the world to create proprietary lock-in. I remember attending a lecture years ago by Tim Berners-Lee with him explaining how the (at the time) academic idea of separating style from content would drive tremendous growth in access and services across the web. He was right, even if we web engineers at the time, were only after better table tags to build slicker sites in Mosaic and Netscape 1.0.

I think, therefore, that the W3C have a pretty good track record in doing what is 'right' and having a true vision of what can be. So what about WHATWG? Whilst I don't trust Apple at all (take a look at what iTunes does as it installs!), I have been a big fan of Firefox and have respect for the work on Opera. (Recent experience with Firefox releases has not been good, however, so maybe there is a change in focus?). My problem with WHATWG is, what's the point?

Briefly the WHATWG version is a little more informal and experimental (and, some might argue more forward-thinking).

Informal? Experimental? To my mind that is a step backwards to the days when we had to extensively code for all the different browsers. The motivation of those involved in WHATWG is not clear, but I would take a guess that it is not as broad or as altruistic as that of the W3C. Indeed, sceptics might just see it as a way to pile a bunch of non-standard code into sexy Apple apps which lock consumers into that platform.

To conclude, I am very wary of this. Banging the standards drum in my product and to my resellers has served their customers well. If there is something positive to come out of it, maybe it will be that the W3C gets some more resources to push their work through at a swifter pace.

[kgun]

Thank you for a thorough and interesting answer.

In this case the W3C appear to be doing a pretty good job. Maybe things don't move fast enough, ...

That may be the reason why WHATWG keep a "living spec".

To conclude, I am very wary of this.

I am not, since Opera is also a participant.

Examples:

http://dev.opera.com/articles/wsc/

http://www.opera.com/company/education/curriculum/

Banging the standards drum in my product and to my resellers has served their customers well. If there is something positive to come out of it, maybe it will be that the W3C gets some more resources to push their work through at a swifter pace.

And the new standard seems more flexible than the earlier specs.

• The word deprecated is no longer used and replaced with the word obsolete for backward compatibility.
• Void tags like <br> need not be closed.
• The HTML5 validator http://validator.w3.org/ is no longer concerned with code style.
• So you can use uppercase, lowercase, omit quotes from attributes, leave tags open etc.
• If you need a stricter HTML5 validater, use http://lint.brihten.com/html/ As an example validate http://thehtml5herald.com/ in both validators and note the difference.

First and foremost I see the new HTML5 standard as more flexible, pragmatic and forward looking. It is still not finished (and will perhaps never be since it is now a living HTML standard).

Note that the "5" has been removed from the WHATWG's living spec http://www.whatwg.org/specs/web-apps...ork/multipage/ that was last updated november 4 2011.

Personally I regard that as an advantage, a true living updated spec.

So who shall or shall not go back too school:

http://www.blognorway.com/2011/06/11...ls-or-w3fools/

[jhannawin]

You're welcome Kgun.

It is an interesting time. I'll spend a little time looking through the Opera stuff with the hope that they keep WHATWG on track.

[kgun]

It is an interesting time.

Do you know http://www.modernizr.com/ ? It will make your sites parse well in older browsers and on other platforms. The JS library is constantly updated and as far as I know it is more targeted at making your sites render well in older browsers. As far as I know from the book I mentioned in the OP it is more efficient on this than the excellent jQuery library.

We know the various WP and Drupal plugins to solve backward compatibility problems with older browsers and other platforms like mobile devices. Now there may be a living JS library that takes care of this for us.

When Modernizr runs, you can for example test if your browser / platform supports geolocation like this:

if (Modernizr.geolocation) {
  // go ahead and use the HTML5 geolocation API.  It is supported
}
else {
  // there is no support for the HTML5 geolocation.
}

View source and look at the head tag of the first link in my signature how I include the library. It should be included in the head.

[weegillis]

@kgun: I'm not sure where I read it... Is it not the goal of WHATWG to hand off their more viable (and stable) specs to the W3C as the final arbiter? This is the way I read it.

In the end, we have a W3C spec (and validator to work with), although for the time we may see a few experimental specs in one version or another of nightly builds in beta versions of browsers. As long as we don't adopt any part of the experimental in our production sites, we should be in the clear.

[kgun]

@kgun: I'm not sure where I read it... Is it not the goal of WHATWG to hand off their more viable (and stable) specs to the W3C as the final arbiter? This is the way I read it.

I am not so very concerned about this. You may know more. To cite from page 4 in the book in my OP:

They (WHATWG) felt that the W3C was ignoring the needs of browser makers and users by focusing on XHTML 2.0, instead of working on a backward-compatible HTML standard. So they went off on their own and developed the Web Apps and Web Forms specifications discussed above, which were then merged into a spec they called HTML5. On seeing this, the W3C eventually gave in and created its own HTML5 specification based on the WHATWG's spec.

And now the 5 have been dropped and it is named a living HTML standard.

Regarding backward compatibility, modernizr (the JavaScript library mentioned above) is as far as I know especially concerned about that.

And it is already integrated in platforms like Drupal:

Drupal integration with the Modernizr JavaScript library. Modernizr allows you to responsibly integrate HTML5 and CSS3 using feature detection, while providing reliable fallback behavior for older browsers.

Source: http://drupal.org/project/modernizr

Example: http://drupal.org/node/1019852

But according to the same book we should not worry. But the living standard may be more up to date and faster to integrate new elements.

[william3964]

I remember attending a lecture years ago by Tim Berners-Lee with him explaining how the (at the time) academic idea of separating style from content would drive tremendous growth in access and services across the web. He was right, even if we web engineers at the time, were only after better table tags to build slicker sites in Mosaic and Netscape 1.0. As long as we don't adopt any part of the experimental in our production sites, we should be in the clear.
---------------------------------------
Travesti

[kgun]

Today web masters should start with a mobile version of their new site if they intend to make it available on mobile platforms. Then making it available for standard browsers using the principle of progressive enhancement. It may be worse to start the other way round. For mobile web applications and web pages, every Kb counts, so minimalism is important. That minimalism may be good as you make your site available for standard browsers.

An example, the well known and highly reputed JavaScript library jQuery has a competitior: Zepto.js

Zepto's primary focus are mobile devices, where small file sizes and tapping into the latest browser features mattergi most for fast loading and optimal runtime performance

.

[ristenk1]

WHATWG was formed by a group of people from Apple, Mozilla and Opera.

Where is Microsoft in all of this? I just had a converversation on a similar note with a coworker the other day; with all of these web oganizations that lay out these standards, how come IE is still the "troublesome" browser. I guess the question isn't if IE can change their browser to fit the standards that the other browsers are coming together to agree upon (because if developers are able to code their content to make it cross-browser compatable then surely IE could just as easily/painfully change their browser), it is why do they not want to work in collaboration?