THANKS TO THE WORK of the WHAT WG, the orations of Steve, the acclaim of developers, and a dash of tasteful pamphleteering, the W3C finally has a hit technology on its hands. Indeed, it has a cluster of hot technologies, the latest incarnation of what we’ve been calling “web standards” since we began fighting for them in 1998, when browser support for HTML, CSS, and JavaScript was inconsistent, incomplete, and incompatible, and the kingmakers of the day couldn’t have cared less. Moreover, after 13 years, the W3C has finally learned that it’s okay to market to your constituents—okay to actively encourage standards adoption.

Hence the HTML5 logo effort, intended as an identity system for all the hot new standards technologies—and initially bogged down by a controversy in our circle about theW3C muddying the waters. The actual muddying began when Steve Jobs announced Apple’s support for HTML5 by pointing to web stuff created with CSS3. In other words, the inaccurate use of “HTML5″ to cover HTML and non-HTML technologies coincided with the surge of interest in those technologies under that inaccurate label. Which is why some thought leaders in our community have reckoned that the business community’s confusion about what HTML5 actually means doesn’t matter so much, as long as they are clamoring for great sites, accessibly designed with web standards—and as long as developers know the difference between HTML5 and, say, CSS3.

In any case, soon after the standards digerati declared the HTML5 banner launch a communications fiasco, it emerged that the launch was actually merely a communications snafu.

An updated FAQ makes it clear that HTML5 means HTML5, that CSS3 is not part of the HTML5 specification, and so on. The W3C’s clarification allows the standards organization to have it both ways in a fashion acceptable to all. In times past, the W3C argued passionately within its own walls during the creation of web standards, only to passively release them as “recommendations” to a world that often ignored them—the development of XHTML 2 in the pure absence of worldly interest was probably the culmination of that phase. But today’s W3C has learned better. It has learned to engage its constituents and to seek approval beyond its immediate constituents—i.e. to reach out to the business community, not just to the authors of O’Reilly and Peachpit books. Its “HTML5″ identity effort represents a reasonable and meritorious effort to cash in on, prolong, and extend the world’s already keen interest in HTML5 and related technologies and practices. Meantime, the little FAQ page and other minor editorial clarifications allow the W3C to pacify its knowledgeable critics and duck the charge that it is blurring the lines between HTML, CSS, and other technologies.

Now that the story appears to be heading purposefully in a single direction, a kink in the works was inevitable.

That kink is also not surprising and not entirely unanticipated. Just when the W3C figures out that HTML5 is hot, the WHAT Working Group (the splinter group that created the actual HTML5 specification in the first place) has decided that HTML is the new HTML5:

  1. The HTML specification will henceforth just be known as “HTML”, with the URL (We will also continue to maintain the Web Applications 1.0 specification that contains HTML and a number of related APIs like Web Storage, Web Workers, and Server-Sent Events.)
  2. The WHATWG HTML spec can now be considered a “living standard”. It’s more mature than any version of the HTML specification to date, so it made no sense for us to keep referring to it as merely a draft. We will no longer be following the “snapshot” model of spec development, with the occasional “call for comments”, “call for implementations”, and so forth.

Those who are surprised should remember that the HTML5 doctype references “HTML” with no version number. In the thinking of its creators, HTML5 was always just HTML. It looked backward (the first web page ever written would be valid HTML5 with the addition of a doctype) and forward. It would continue to evolve. The WHAT WG gave itself the job of steering and updating HTML, while the W3C took on the task of maintaining milestones (a task it will continue to perform).

In practice, the WHATWG has basically been operating like this for years, and indeed we were going to change the name last year but ended up deciding to wait a bit since people still used the term “HTML5″ a lot. However, the term is now basically being used to mean anything Web-standards-related, so it’s time to move on!

To those inside the circle of trust, there is no contradiction here. The W3C will doubtless continue to market HTML5, and, for a time, design technologists will continue to write HTML5 books and teach HTML5 classes, if only to acknowledge HTML’s new capabilities and to clearly mark the break from the technologies and practices of the past. Eventually, quite probably, the WHAT WG’s view will take hold, and we will view HTML as a living specification.

Meantime, we’ll take 5.

Thanks to J. David Eisenberg for the nudge.

89 thoughts on “HTML5 vs. HTML

  1. “HTML5 will be very exciting in 2011, because anything exciting in 2011 will be called HTML5.”

    I suspect inner circle standardista posturing and pamphleteering makes more or less no difference to the broader business community’s interest in, and adoption of, web technology in general.

    And on mobile (where I think it really matters), we should take HTML-anything over pointlessly-native apps.

  2. Yeah, I have to say: HTML doesn’t need to be tied to a version number. It never has, although people have tried to impose a meaningful versioning system by arguing that doctypes actually mean something.

    By its very nature it doesn’t require any kind of version number or versioning mechanism. The WHATWG is only acknowledging this.

  3. But html5 is the goto web technology for people who want to root for something other than flash, not html. Html is old and useless nowadays. Right guys

  4. I’m just trying to decide what to put on my resume. If i just put HTML, should I trust the employer is hip to this? or should I put HTML, XHTML, HTML5 ?

  5. I actually prefer the more futuristic term “HTML5000″ for the umbrella of HTML plus CSS3 plus all the new standards. It sounds like you’re building web sites from the future!

  6. A well written, and well reasoned summary, sir. Personally, I’ve seen the divide between the WHAT WG and W3C as a point of confusion for many rookie webbies for quite some time. Hopefully their roles will become clearer to those outside the innermost circles of specification land as a result of this.

    Moving on… and the confusion sparked by the fruit company not withstanding, using HTML5 as a marketing term is fine with me. In fact, I think it’s brilliant. Now that Jeremy Keith, et al. have had their way with the FAQ (and rightly so), we advocates of web standards should applaud and reinforce the efforts of the W3C. If we’re to be effective in our advocacy in the face of the savvy marketing of products like Flash, and the general ignorance of modern web technologies, we need a sharp edge on our blade.

    Having an umbrella term helps focus attention and generate excitement. “AJAX”, “Web 2.0″, and *cough* both “SEO” and “social networking” *cough* all being perfect examples of this. (Mr. Croft pointed this out, too). No war was won without a rally cry. It’s time we stopped debating and berating each other over these semantics in front of the whole class and speak with a unified voice in support of HTML5 – even if it’s not exactly the right term.

  7. You gotta give Apple and Steve Jobs some credit for html5 getting so popular. Although Jobs “muddied the waters”, his “Thoughs on Flash” letter proclaimed that the web should be open, not based on third-party plugins, and that’s why “Apple has adopted html5″. The fact that people want their websites not only compatible but reaching their highest potential on Apple mobile devices is the driving force behind html5 becoming so popular, if not a key player.

  8. I think ‘HTML5′ was the battle that needed winning in order to reset our understanding as to what ‘HTML’ means. Had the initial push for new HTML standards just been named ‘HTML’ to begin with, I think it would have caused confusion and misconceptions of moving backwards instead of forwards.

    Now that HTML5 has become accepted and studied by designers everywhere, the understanding that it really is just HTML has spread. This makes the transition to just calling it HTML much easier.

  9. @Tom Wolber:

    ’Twere me, I’d put

    HTML (2.0, 3.2, 4.01, 5)

    If the cluefulness of the evaluator seemed to be at issue. But only if. They’re bound to ask, you just want to get your cv pulled from their crappy database, if it gets pulled from that (as opposed to an e-mail inbox alone).

  10. I like the idea of just calling it “HTML”, which is an evolving standard, with many features, with varying levels of browser support at any given time.

    It’s always seemed odd to me to label the big collection of features as “HTML5″ since it really needs to be considered as individual features, a few of which are useful now, and other that will gradually become more useful over a period of a decade or so. (How long will it take until IE9 is the oldest version of IE that you care about?)

    When would you cut off changes and call it HTML6?

    Another issue we see all the time now is clients coming to designers and saying “my site needs to be HTML5″, even though they really have no idea what that means or what the implications are. Better for them to focus on capabilities — “I want my video to play on iPads and iPhones”.

  11. If you recall, the W3C tried to brand these technologies the “Open Web Platform“, but it never took off, HTML5 did instead.

    With the removal of versioning from HTML, HTML5 can take center stage as the brand for the full web app stack that the W3C is pushing.

    Then next step is to develop a that is articulate and holistic about what the Open Web Platform is.

    It’s kind of funny; the debate that happened in the comments of this and Jeff Croft’s blog back in August 2010 has turned on it’s head. Whereas then, the the debate was whether or not HTML5 was becoming a brand that lumped too many technologies together, now it’s a debate on whether or not it lumps enough of them together.

  12. All I know is that there will be no Flash or Silverlight in my development future. And no native apps either. The only thorn in my ass is Internet Explorer. I hope Microsoft gets their act together! And this is coming from a .Net guy!

    They can call it whatever they want. I’m having a blast with HTML whatever, CSS and JQuery.

    Paul Speranza

  13. @Paul Speranza: Wow, a .Net guy into HTML5! It is possible! (joke)
    With IE9 even big slow Microsoft realizes that web standards are the only way to go. So, sorry to say, but every developer or designer not paying enough attention to the HTML code they produce, is putting their reputation in serious danger.
    I think the following thing we should strive for is kicking Word out of business. Every digital text should be put default in HTML.
    HTML is, was and will always be the only true foundation of all things web. Kids should learn HTML not Word.

  14. I don’t think I buy Jeffrey’s argument that the FAQ clears up the confusion for developers. But I do agree that we need a buzzword.

    In an ideal world, I’d have made that buzzword “web apps” (or similar)—i.e. a broad term which the public could understand, but which doesn’t reference specific specifications.

    Of course, we developers don’t get to choose the buzzwords that catch on; Apple picked “HTML5″ so HTML5 it is.

  15. After reading about HTML5 as being just the continuation of HTML 4, and with all the buzz around it, I didn’t really get why it’s HTML5 instead of HTML 5. So, now, HTML it is. HTML for web-app steroid!

  16. “…today’s W3C has learned better. It has learned to engage its constituents and to seek approval beyond its immediate constituents—i.e. to reach out to the business community, not just to the authors of O’Reilly and Peachpit books. ”

    THAT gives me hope for HTML5 remaining pure. That they’re educating the early adopters as quickly and efficiently as they can.

  17. I see people turning a blind eye to the problems that have happened, and will continue to happen, now and into the future, because of the unresolved conflicts and confusion about the W3C and WhatWG. I believe many do so because they want to be seen as uber hip, as they knowingly nod to each other, and make proclamations that the problem is really an issue of “confusion among newbies”, or some such blather.

    Consider this: there is no WhatWG. There is no organization that is WhatWG. WhatWG is Ian Hickson. WhatWG is hosted on a web site owned by Ian Hickson, maintained by Ian Hickson, and controlled by Ian Hickson.

    So let’s not look directly at the dangers inherent with allowing one single person to have almost unlimited control over HTML. Let’s not look at the haphazard way the specification is hacked together, and suddenly grows and shrinks depending on nothing more than whim and fancy, because everyone is desperately trying to stay relevant with this uncertain, and grossly mismanaged, specification.

    Well, peachy keen. Have a nice day.

  18. What they should do is call all this standard-stuff-in-yer-source something fancy. Call it “TheCode”. All things are either following “TheCode”, or not compliant. Thus, HTML, CSS, JS, whatever you want to circle the wagons around, are the MINIMUM. It’s the Acid# test. Anything beyond “TheCode” is simply proprietary. Yes, “TheCode” will evolve, but dammit, give us specs we can work under.

    If we know something is going to have a 5 year shelf life, we can hack the hell out of it, stay compliant, and get consistency. This is the moment you tame the Wild West, Working Group.

  19. The big problem with versioning HTML, I think, is that it actually never served any purpose, apart from getting you out of quirks mode.

    It’s not like I can tell a browser what stylesheet to load depending on what version it supports, so actually including it as metadata or whatever would be useless. You’d need to still care about what features are supported in each individual browser at any given time.

    If you could rely on version numbers to actually guarantee that all the features inside that spec would work, then it would be another story. Maybe then creating more version numbers with smaller amounts of new features would be useful.

    But for that to happen, either you create a platform that is independent of the browser makers, or you need to get them to agree on what get’s implemented and act in synchrony. The latter is just fundamentally impossible, because they are competing at the rendering engine level. In their mind, having new features that no other browser has is a good thing, but actually nobody uses them until enough browsers support them, hence not really a competitive advantage apart from throwing marketing messages around.

    They are reinventing the wheel, actually. The rendering engine is the piece of a browser we actually want to NOT be different from one to another. A good start would be every browser using WebKit, and then contributing to the project any innovations they make on the rendering (the chrome, history and bookmark management, etc, would still be different).

  20. What is this habit of “looking backwards”.

    You say the “first web page ever written would be valid HTML5 with the addition of a doctype)” but of course this is not true, because blink tags and font tags, which were common during the first years of the web, are not valid at all today.

    So why even bother to look backwards? Who cares about old people who don’t follow web standards. Leave them in the dust!

    I would think that instead of backwards compatibility, the motto of the W3c would be to skate to the where the puck is going, not where the puck has been.

  21. Why we should continue tu say it is “Okay” all the new web technologies are marketed as “HTML5″ ? There is no other professional field where things are called by other names. Imagine something like that in medecine or philosophy…it is not okay it’s that we are still an unformed professional field, and whe shouldn’t be “okay with that” if we want to mature and acquire a solid status. It is not about semantic or marketing or pragmatism or standardistas deviations. It is unprofessional to name a specific technology, or any other piece of knowlede with an inaccurate name. Isn’t it?

  22. Miro, overall I think you have a good point. However, you state:

    The rendering engine is the piece of a browser we actually want to NOT be different from one to another.

    But that’s only true if you’re only interested in html & css. If you care about behavior, you also want the implementations of javascript, the DOM, services like geolocation & data storage, etc. to be the same from browser to browser as well. Standardizing on a rendering engine would be nice, but it would not resolve all the things that are important to standardize.

  23. Google Chrome has shown us the versioning is silly when the “user” is not required to update. I use Chrome. Where is if I were using Internet Explorer I would say, I use IE8.

    With HTML, the user simply gets the latest version of HTML that the developer has provided. Other than for developers working on the draft, the versioning is irrelevant.

    That being said, thank god for the HTML5 hype. People buy hype, so if it takes HTML5xs to keep the dollars rolling towards web standards, fine by me.

    @Tom It depends. If you are confident you will get an interview, list HTML. Then during the interview you can mention HTML5 elements and such, as laugh off how you find it funny people getting caught up in the numbers. If you are concerned about landing the interview, I’d mark your resume with HTML5.

    @Paul S – I felt the same. Wrote off Flash….well until Google said goodbye to H264 in favor of WebM. To push video to Chrome and your iOS viewers Flash may yet have a longer future than we had hoped.

    @Charra “Blink tags, font tags” and such where not part of the “very first web page written”. I think you may have misread his statement to be all webpages are valid html5. As for the rest of your comments… well I’m just very happy you have found for what I assume to be the first time. The more you read here, and the more you watch on shows such as (some of which zeldman hosts :) the more answers you will find to your questions posed.

  24. @Christopher Beckwith

    Sorry, but Chrome became irrelevant with that bonehead move from Google to ditch h.264.

  25. Yeah, Web development needed a rebranding. That is the key use of the term “HTML5”.

    People who aren’t developers think Web development is making content to run in IE6/FlashPlayer. We have some mobile device makers shipping an elderly WebKit that gets on 91% on Acid, but because there is an optional beta install of FlashPlayer for Mobiles they are calling that “the full Web.” That is part of the profound failure of the XHTML era. Consumers routinely had 2 or more Web browsers and routinely could not see audio, video, or other content within pages. Total failure. We need a way to say “the open, device-independent, consumer-friendly, post-XHTML, post-IE6, post-FlashPlayer, post-proprietary, post-PC Web,” and “HTML5” is a lot easier to say.

    Part of the reason the CD worked was branding. You got a player that had a CD logo on it and content that had a CD logo on it. That is what is needed for consumers, because it enabled them to purchase players and content from any manufacturer or publisher with the assurance it would work. And publishers could create and share one CD and be assured their audience was listening. Web browser makers and publishers need to enable consumers to choose an HTML5 browser and HTML5 content for the same reason, so that their freedom to use any brand or class of of Web browsing device and view any web content is not compromised. And Web publishers also need to be able to deploy one “HTML5 app” for everyone.

    If you are a Standardista who chafes at the HTML5 versus HTML naming then you should consider that part of your penance for XHTML. Complaining only draws attention yet again to the trillions of dollars of damage and billions of wasted work hours caused by the failure of XHTML. Even now, when I say “XHTML” I spit afterwards to get the taste of failure out of my mouth.

    > on my resume
    > should I put HTML, XHTML, HTML5

    Just HTML5, because it includes HTML (as well as CSS3, SVG, etc.) and obsoletes XHTML.

  26. I wouldn’t trust the W3C to put together a standards document to take out the trash for weekly pickup. When Hixie goes around cavalierly throwing around like 2022, he winds up giving the Flash apologist camp the coup they are looking for: For all their good work, sometimes the standardistas are their own worst enemies.

    The WHATWG dropping the “5” from HTML5 was stroke of genius, and shows they at least understand the wider world we all live in.

  27. Miro – I recall people making the same comment over Mozilla’s Gecko engine, and the waste of effort Apple were making in investing energy into WebKit that could have gone into the One True Rendering Engine.

    Instead, what we’ve seen is a really exciting era where Mozilla, WebKit and Opera (at least) have competed, in the sporting or scientific sense, to meet and beat various benchmarks, and to implement new features.

    I’d also say that a large part of the history of the development on HTML as a standard IS around proprietary features being implemented, then standardised – such as the DOM (Netscape), XMLHttpRequest (Microsoft), Canvas (Apple).

    In fact as we know from the bad old days of 95% IE dominance, what is in the spec is irrelevant to developers – we have to develop to the implementation(s) in use.

  28. I agree wholeheartedly with Shelley’s earlier comments about WhatWG and Ian “Hixie” Hickson:

    Consider this: there is no WhatWG. There is no organization that is WhatWG. WhatWG is Ian Hickson. WhatWG is hosted on a web site owned by Ian Hickson, maintained by Ian Hickson, and controlled by Ian Hickson.

    So let’s not look directly at the dangers inherent with allowing one single person to have almost unlimited control over HTML. Let’s not look at the haphazard way the specification is hacked together, and suddenly grows and shrinks depending on nothing more than whim and fancy, because everyone is desperately trying to stay relevant with this uncertain, and grossly mismanaged, specification.

    The IETF hybi working group interfaced with WhatWG and Hixie, who edited the websocket specification as an internet draft. That draft went though 76 public iterations — it’s unusual for an internet-draft to have more than ten public releases after discussion, at a typical rate of once every few months. But that reflects how Hixie chose to work. It became impossible to tell the minor draft modifications from the major changes, as he simply published his workflow. After criticism as to which version of the draft implementers should work to, minor vs major changes, comments on the draft not being addressed, WhatWG being treated as equal to an IETF group which made no sense for an IETF specification as an output of the group, etc., Hixie handed control of the websocket draft over — to another Google employee.

    WhatWG saying ‘there is no HTML5, just HTML’ is a reflection of how Hixie likes to work – continuously, and without milestones or deliverables which could be criticized. Interoperability testing between browsers will become a nightmare as the specification is a moving target. (Is there an interoperability problem, or did they just work to slightly different versions of the specifcation?) Major changes will not be visible in the plethora of minor updates. The amount of attention that is required by all to keep track of things and decide whether they’re important or not, and whether your codebase should be revised to match, will increase dramatically.

    When Hixie says ‘we, read I. When Hixie writes:

    We will no longer be following the “snapshot” model of spec development, with the occasional “call for comments”, “call for implementations”, and so forth.

    he means that opportunities to comment on his work at clear intervals are not welcome. It’s job security for Hixie. Grossly mismanaged? Oh yes.

  29. @Matt Ryan – You’re totally right. All those things should work exactly the same across all browsers. I called it rendering engine, but I actually meant all of the things you mention too.

    @JulesLT – And I think they were right. Just imagine how fast JS engines would be if all the effort invested in them had been made into one implementation… On the other hand, the Mozilla guys would have needed to accept input from outsiders to their organization, and I don’t know if they were even open to that. I’m not sure if Gecko would have been a great “One True Rendering Engine”…

    Yes, we’ve seen an exciting era, but more in the sense of spectators of a sports game, where you watch how they compete and maybe root for one or the other. But we don’t know how it would look today if instead of competing they had collaborated. Browser fragmentation is still one of the biggest issues that a developer has to confront with, usually resorting to the lowest common denominator, instead of using the latest features because, obviously, they would break in non-supporting browsers.

    And again, you’re right about the history of development of HTML. What the standards organizations have been doing is just putting on paper what is being done by the implementors, not the other way around. The one time they tried (xHTML2) they failed big time… It just goes to show how dependent they are on browser makers. They have no authority at all, they just “suggest”, if anything…

  30. @Eric,

    Sorry, but Chrome became irrelevant with that bonehead move from Google to ditch h.264.

    Comments like this show a lack of understanding of the very reasons for the existence of standards. If a browser exists and is standards compliant and is fairly widely used then you can’t make a decision about whether it is relevant or not, unless of course you are developing web applications for yourself to use.

    H.264 is not a web standard, the tag is. h.264 is a patent protected technology.

    In the real world developers and producers of web content and applications are trying to reach as wide an audience as possible using the least resources in terms of development time, lines of code, bandwidth and server and client processor cycles. Competing browsers, mobile platforms, and more importantly user expectations have to be catered for – otherwise – you’re irrelevant.

    @Jeffrey Zeldman why flame Charra? The first HTML doc wasn’t HTML5, HTML5 doesn’t really exist right? HTML5 is just HTML. HTML is HTML is HTML, or just hyper text markup language, this isn’t getting us anywhere…

    It’s no wonder that the industry, clients and “newbies” are confused. Less the semantic web, more a tedious argument over semantics.

    Competing authorities on web standards – now that really does confuse things in the standards “community”, in the real world, more so.

  31. @Magnus
    Perhaps “newbies” don’t understand everything, but this newbie understands one thing:
    “standard”, “open” and “patent protected technology” are NOT mutually exclusive terms.

    Don’t be a bonehead by bringing your ideology into it.

  32. @Kizedek –

    Don’t be a bonehead by bringing your ideology into it.

    Whoa! It’s nothing to do with ideology. I don’t care who makes money, Google, Apple, Adobe, it’s all the same to me. You shouldn’t care either.

    I’m just glad that the standards don’t dictate which patent protected technology I need to adopt to build standards compliant applications.

    I’m not a bonehead. I was just pointing out that the h.264 codec isn’t really part of the the updated HTML standard. You are the bonehead – if you’re a newbie you shouldn’t be throwing around insults so readily.

    You could have just flamed me for failing to close my blockquote tag but instead you insulted me with a very weak point. You understand one thing, give yourself a big pat on the back, well done.

    Sadly it appears that “newbie” and “opinionated” aren’t mutually exclusive terms.

    Don’t tell me what to say, your point without resorting to insults, would have been fine (still inconsequential and irrelevant none the less).

  33. OK, so have we all agreed on how many angels CAN dance on the head of a pin?

    Well, while you guys all fight about it, I’m going to go pay seven bucks for a cup of coffee with air bubbles in it.

  34. That “links and anchors” page is just a demo page. IT’s not a real website (in the sense that no one uses it). Real websites back then did have blink tags and font tags, which are not valid anymore.

  35. Charra, you’re switching from apples to oranges. I said the first web page would be valid HTML5. You referred to blink and font tags. I said the first web page didn’t have blink and font tags, and showed you the first web page. Now you say the first web page doesn’t count as the first web page because it didn’t use blink tags and font tags. Circular reasoning isn’t reasoning at all. It proves nothing.

    Also. You say the first websites used blink tags and font tags, but that’s also untrue. I worked on one Warner Bros’ first websites back in 1995. It had no blink tags and no font tags. They didn’t exist yet. But it was a real site, with 1.5 million viewers at a time when the entire web-using population was 3 million. In other words, it was a real site that was viewed by 50% of the web-using population. Only Google and Facebook can approach that kind of traffic today.

    Anyway, not to pick on you, guy, I’m just saying, it’s best not to pick a fight over details when your knowledge of those details is wrong.

  36. Shelly, do you really think people are using HTML5 because they want to seem hip? You don’t think new publishing-oriented semantics and features like video and canvas might have something to do with it?

    I’m not disagreeing with your concern about so much power being concentrated in the hands of one Google employee—I’ve worried about that here and elsewhere from time to time. I came down on the side of HTML5 because I think it has good things in it, and because the alternative was XHTML 2.

  37. The argument that Chrome proved versioning to be silly is great, but misses on giant point…Chrome automatically updates, whether you like it or not. Therefore you ALWAYS have the most recent version. As @zeldman just said, apples and oranges. It kinda makes me think that all browser vendors did this.

  38. I ordered one of these new HTML5 t-shirt not because “HTML5 is the future” but because it’s a good looking t-shirt. HTML is the future and it will be easier to evolve if enhance only bits of it instead of pushing for version number updates.

    From a developers perspective it will be great to incorporate new parts of HTML into your website if you knew that X, Y and Z browser had support for it. On the other side it will be a lot easier to market news in a new version of a browser if “Browser XX supports this new thing” in HTML.

  39. Mostly I think the problem is not the lack of space before the “5” but that it is not in parity:


    When used in parity the “5” can take center stage and now can be used as a versioning acronym – “5, 3, 9″ (in this case html5, css3 on IE9), etc..

  40. I’m with Lloyd 100% on the Hixie aspect of HTML5… it has long irked me that this is (not even secretly) a one man band called a public specification. This has been my concern from the beginning. My concerns increased when I started to get in conversations with high profile web standardistas who attend conferences who were afraid to mention Hixie in a conversation unless he were to find out about it… and what? Has someone become that powerful as a person that we should be afraid of their wrath and effect on our career?

    However, HTML5 was the only alternative to XHTML2, which was going nowhere. So as it stands it’s the only option that provides improvement. HTML5 is the way forward. But having said that I would add that by no means is it a sure thing that the road forward won’t be full of land mines simply because we’ve been seduced (or told) to expect everything for everybody in the specification.

    Can you name one specification that is never a finished document that is instead a “live document”? That can never be relied on from day to day to remain the same? How do you have a conversation about the “current” specification? How do you agree that you both read the same specification? It defies what I understand as software engineering as an undergrad, at least.

    I think lack of versioning will bite us in the arse… and the potential danger of this path in the wild is that in a very few short years everybody is working to the different versions of a wild specification that nobody but Hixie still understands… because to understand it you would have to keep reading and understanding every revision on practically every single day. And I’m confused why that is a surprise to some people. A standard is a set bar that everyone understands as correct usage… a non-standard is having no set bar to tell everyone correct usage. The live document, to my mind, smacks of web non-standards. Not now, of course, because it looks peachy now… come back in 5 years and tell me the state of HTML5 without versions or a set specification.

    So I agree with Lloyd… and I think that Hixie can out talk, out smart and out politic nearly everyone in town. He could pull in a band of illiterate gypsies and a crystal ball and convince the majority of this industry that he’s got all the answers.

    So at this stage HTML5 yes… because there is no alternative… but versionless HTML no… because it makes zero sense in the real world. And it makes even less sense in a world or context of machines talking to machines. Apologies for the rant… time for the first coffee of the day.

  41. @Shelley You’re right, the WHATWG is controlled by Ian Hickson — despite all the input from implementors and authors, the final say is his. However, I’m almost certain that I read somewhere (perhaps in Jeremy Keith’s book?) that Hixie can be sort of vetoed by other members if he tries to do something mad. Am I right on that? (sorry, didn’t have time to check)

    Besides, wouldn’t you say the real power is in the hands of the browser vendors? I mean, if Hixie writes something into the spec, mad or not, and the decision makers for Microsift, Mozilla and Webkit say “no” then that’s that — the spec is just fiction.

    I am writing web pages in HTML5, to use its former name, but not because I think it makes me seem hip. I haven’t really networked and I am the only web person at my company, so I have no one to impress with my hipness. I do it because I think it’s better.

  42. I have let go of chasing HTML5, I thought it was not mature yet, then Chrome dropping H.264 I thought may set back who agrees on what … zzzzz … basically just waiting for things to settle, this article brings me up to speed on many things, like I didn’t even know the new HTML existed. Thanks JZ for a clear summary of what is going down, wait Jay Z? (nerd-giggle-snort)

  43. @Lloyd: If you need an opportunity to comment on the HTML spec at clear intervals, you are more than welcome to do so on the first of each month. We can call it Lloyd Commenting Day if that would help. Personally I would be happy to receive your comments any day of the month.

    @David: Yeah, the charter members can overrule me whenever. In practice they won’t need to, though, because I want the work I do to be relevant, and so the specs I write will match what the browsers do — so if the browsers disagree with something I put in the spec, they’ll just ignore me and do their own thing, and I’ll have to change the spec to match.

    Ain’t much power in the spec editing job. Either you do what the implementors want, or you become irrelevant.

  44. This post is good, well-reasoned thought from a voice that commands respect. I’ve been a little puzzled why this subject has been eliciting so much derison from ordinarily reasonable (I assume) people. So, it’s refreshing to see a voice of reason arguing that “clarity,” at least from a larger perspective, may not even matter.

    If clarity is going to live in the eye of the beholder anyway, then empower now that which makes the most sense for the most people.

  45. “If you are a Standardista who chafes at the HTML5 versus HTML naming then you should consider that part of your penance for XHTML. Complaining only draws attention yet again to the trillions of dollars of damage and billions of wasted work hours caused by the failure of XHTML.”

    Hyperbole, much?

    David Gross, I know of one specific instance where a so-called member of the WhatWG inner circle did object, and was ignored. The person has not participated in the WhatWG email lists, since. In fact, half the people from the original WhatWG inner circle no longer participate.

    This was a self-invited group that originated 7 years ago to “clean up forms”. It has morphed into basically one person’s ownership of HTML. And I’m absolutely astonished at the web community’s complacency in this regard.

    Jeffry, the vast majority of web developers and designers will never have a need for video, or canvas. The vast majority of web pages need neither.

    (Speaking of which, Canvas was around before HTML5. It was implemented in most browsers before HTML5. What the W3C provided–not the WhatWG–is a patent policy that made Apple comfortable enough to contribute Canvas. )

    But to answer your question, yes: most uses of HTML5 today are because people want to use the newest, the coolest, the hippest, the edgiest of the technologies, whether they need it or not. That’s one of the things that a version provides–not only does it allow people to identify the version as part of the constraints for the project, it also pulls people along who might otherwise just leave their pages, as is. Versions generate excitement.

    But leaving aside the foolishness of “living standards”, my concern is less on why people use HTML5 and more on the current state of HTML ownership.

    WhatWG is owned and controlled by one individual. People can pretend this isn’t so, and participate in a tech version of the King Has No Clothes, but this is a fact that cannot be denied.

    The reality is that the web community is allowing one individual to have too much control because it’s caught up in all of the hip uberness; the browser companies are abetting this, because it’s to their advantage to have things in this state; the W3C is going along for the ride because it’s desperately trying to find a way to maintain some semblance of standardization of these critical web technologies, in the face of browser companies determination to undermine the standards process.

  46. I have said it before and I am going to say it again. HTML5 is not compatible with earlier versions of HTML. It does not matter how many times that we say it is backwards compatible when it isn’t. Why not? Because it redefines the semantics of existing elements like small to something that is semantically incompatible with the semantics it had in previous HTML versions (the small element is in HTML 3.2). Because this issue is about semantics and not syntax a validator cannot help you to find this issue and will gladly report the document as valid anyway. This could be fixed quite easily, for example by creating a new sidecomment element instead of changing the semantics of an existing element, but so far no one has any interested in doing so. This is just one of several reasons why using the HTML name to describe HTML5 just add confusion.

    All this focus on the cool new features like canvas and video have taken away focus from other, maybe not as cool, issues like semantics and accessibility. Ian Hickson claims that their fork of HTML5 is “It’s more mature than any version of the HTML specification to date” but if that is the case why is there a large list of open issues that needs to be solved before HTML5 moves the next step in the standardization process? Is it because these open issues are not cool enough because many of them relate to accessibility?

    If everyone just focused on getting those issues closed instead of adding confusion by trying to rename HTML5 it would be far better for the web in my opinion. Once HTML5 has become a standard then people can start inventing crazy cool things for the next version of HTML.

  47. Hrm.

    As a software dev, “HTML” concerns me greatly. To me, it’s yet another thing that will slow down the progress of web standards.

    Why? Simple – there’s an axiom in the software engineering world along the lines of “Avoid big design up front”. Trying to design something absolutely right the first time is pointless and silly – the better option is to give yourself an easy out later on by making yourself more resilient to change (refactoring).

    By unversioning HTML, the WHAT WG are to some degree removing their ability to *fix* broken parts of the standard. Lets say in 3 years time we all decide (unlikely, but hypothetical) that et al are might stupid ideas and want to stop people using them in favor of something else.

    Well, what is a browser developer to do? They can’t optimise their engine for the latest standard. Instead they have to do exactly what they do now – provide thousands of backwards compatible hooks for old versions of HTML.

    Personally, I don’t buy the argument that it’s more work to keep incrementing the version number because manufacturers will have to keep adding on rendering pathways. Just deprecate anything older than x versions to baseline rendering only and be done with it!

  48. I’ve been an early adopter of HTML5 (end of 2009) and have just been irked by the lumping together of CSS3 (I know they backtracked on that one, sort of…) and almost anything else under the HTML5 umbrella. It confuses things.

    Bruce Lawson has come up with the term NEWT and we should all use that.

  49. Steven Clark said

    “Can you name one specification that is never a finished document that is instead a “live document”? That can never be relied on from day to day to remain the same? How do you have a conversation about the “current” specification?”

    This troubles him, as it “defies what I understand as software engineering as an undergrad”.

    I was an English Language and Literature undergraduate, and it doesn’t trouble me. An analogy I’ve used before is that of the English Language. To paraphrase Steven – “How do you have a conversation when the English Language isn’t even finished yet?”

    Like using CSS, we choose the bits that are understood widely. Some parts of CSS 2,1 (itself “not finished”) are implemented nowhere, so they’re not ready for use. Parts of CSS 3, like border-radius, are mature enough to use everywhere without vendor prefixes. This doesn’t trouble us. See my joking post from October “CSS 2.1 “not ready for use” says journalist”

    Hickson said it’s the browser makers who control the spec not him. It isn’t. Ultimately, t’s the development community. The browser vendors were scared by things like Flash and Silverlight and the way they were meeting developer demand for widgets and whistles for application development, so set to work on a spec that provided an non-proprietary alternative.

    If only the development community would be content with static documents and blue underlined links, the browsers wouldn’t need to keep innovating, everything would be lovely and there would be no uncertainty. Which would be much better for all, wouldn’t it?

  50. Funny how the same people who only a few months ago cried “HTML5 is a brand that includes everything and our dogs, never mind the confusion”, jumped when W3C did the same thing. And while I completely agree that for once they were right to jump, to me it seems that they did so for the wrong reasons. It was not semantics that made them react, but the voice of a few well-known names in the field that, thankfully, rang the bell – so they rallied behind those voices, most of them without advancing much into the “what” and the “why”.
    But now the same crowd, instead of focusing on advancing the new technologies, finds itself at a loss – they lack a term that would encompass said new web technologies, a term that they could use like a badge of honor, something new to market themselves with. Now NEWT sounds like something from Harry Potter and it doesn’t quite carry the weight that “Web 2.0″ had. I was thinking of suggesting “Duckie”, but then people might not take one seriously with that and it would be quite difficult to market. I am looking forward to new additions to the list.

  51. We discussed this issue here last year. I prefer “HTML5 and related technologies.” We don’t need yet another buzzword, and we shouldn’t ignore HTML5’s momentum and traction in the business community. Don’t reinvent the wheel, don’t mistake the wheel for the car, and don’t call a wheel a triangle because some people think wheel = car.

    God, I’m good.

  52. WhatWG is owned and controlled by one individual. People can pretend this isn’t so, and participate in a tech version of the King Has No Clothes, but this is a fact that cannot be denied.


    I don’t see that having one person in charge is necessarily bad.

    CSS1 was created by two people, Bert and Hakon, and that worked out okay.

    CSS3 is being developed by lots of more “democratic” groups working in parallel on separate modules, and it’s a bit of a clusterfuck.

    XHTML2 was essentially driven by one guy, who didn’t want to hear what he didn’t want to hear, and that didn’t work out so well.

    What do these examples show? They show that the quality of a standard has little to do with how many people are involved in its creation.

    Hixie is just one guy, and the future of HTML is a lot of responsibility for one guy. I’m with you that it’s a lot of power in the hands of one person—and one person who works at a very, very powerful company with its own agenda to advance. But Hixie listens to feedback, and the feedback-giving and decision making are taking place publicly, not in secrecy.

    Hixie is not the Mubarak of HTML. And if he were, to whom should he cede power? Is there a group of people who would do a better job of steering the spec than he is doing? It seems to me that outsiders are contributing to the shape of HTML5—Hell, I’ve done so myself—and Hixie is supervising. Call him the Scorsese of HTML.

    What’s your alternative?

  53. Damn me, I should have inserted some sarcasm tags. Le sigh.
    But to expand upon it, I think that Bruce is right, the developers are the ones who push new specifications into happening and also the ones who market said specifications, correctly or not. The problem is that if things were clearer from the beginning and people wouldn’t choose to misuse terms just for the hype of it, the web community would have better chances at understanding, advancing new specifications and making room for future ones. Instead of focusing on buzzwords, that is. I’m rather sure that my code won’t freeze without them.

  54. What appears to be unrealised in this discussion is the role of the W3C HTML working group in the continuing development of HTML. Due to the W3C’s involvement Hixie is neither the Scorsesee or Mubarak of HTML. Granted he plays a very important role, but he is not the ‘supervisor’ of HTML or HTML5 and the future of HTML is not ‘his responsibility’. Where I come from (Australia) we have a bicameral system of government, the lower house where policies are introduced and debated and the upper house where those policies are reviewed. The WHATWG can be likened to the lower house and the W3C to the house of review. If there is disagreement about Hixie has put into the spec the W3C provides the process for review and change. In this process Hixie does not hold the power, nor do the browser vendors, they have a seat at the table along with the many other stakeholders in this thing we call HTML. The W3C provides a process for review, the WHATWG does not.

    Whether one likes or agrees with the never ending standard (Hixie) or the iterative approach (W3C), it is the web that would be worse off either one were to disappear or the importance of either in the continuing development of HTML were to be dismissed.

  55. @bruce: I understand why you use English as an example of an unfinished spec, but I’m not sure I buy the argument. In order to have conversations about “proper English,” whether grammar or spelling, we rely on some form of milestone. Spelling and definitions of words are generally defined by dictionaries, with new words being added and old ones being dropped with each new edition. Grammar and the like are similarly defined in books like the ALA Style Guide. They reflect the current “best practices” with regard to sentence construction. Together, these resources serve as the milestones of the English language, showing where it is/was at a given point in time.

    I do see value in having versions of HTML (and other languages), not only to be able to track their evolution (HTML3, for instance, was an amazing spec, but never saw the light of day), but to establish a baseline against which compliance can be tested. I’m fine with the WHATWG’s decision to drop the 5 from their work on the HTML spec; that makes sense. What I do want, however, is for the W3C to keep creating the milestones against which HTML authored (by us) or implemented (by browsers) can be compared. There is great value in being able to see where we (or they) stack up and where improvement is needed.

    Going back to the English language for a moment (at least the American variant), you can see this in play on the Merriam-Webster site: users submit new words for consideration and the company (and their panel of linguistic experts) make the call as to when words should be added to the English lexicon as part of a new edition of their dictionary. Easy peasy.

    PS – @Jeffrey: I could be wrong, but I think you intended “W3C muddying the waters” to point to this WaSP Buzz, not to be a second link to Jeremy’s book.

  56. If anyone cares or makes it down to the last comments…

    1. There is a difference between infrastructure and applications.

    Consider IEEE802.11n. If that spec had been a “living standard” we would be in for a nightmare. Or LTE or plain and simple Ethernet. Stability is the very key issue that makes these standards work.

    The same goes for IP, TCP, UDP, ICMP, DNS, DHCP, etc.

    Simply put, the further down you are in the OSI model, the more stability is required from the spec. And until a spec reaches its final state any implementation is considered an experiment and buyers (outside of the ignorant private consumer section) beware and do not buy any stuff.

    However, at the top of the OSI model we have stuff that works really well, that are completely un-specced – like PHP that drives Facebook and Wikipedia. The implementation is the spec, and the implementation process is chaotic, to outsiders at least.

    HTML has more in common with PHP than it has with WLAN technologies. But let’s for a moment consider how they work:

    Manufacturers get together to form a working group, under the umbrella of a parent organization, such as IEEE, EIA or IETF. They iterate around a number of RFC’s, managed by a spec editor, until they have reached an agreement. During that process they develop chipsets, firmware and prototype products in their lab. Thus the chaos is invisible to ordinary men. However, this sounds a lot like the WhatWG core idea. Except for the fact that there are versions of the RFC’s.

    But as long as the W3C is kept in the loop, there will be versions of HTML. And versions are a good thing when we establish baselines of support. When we have tests that check for the quality of an implementation, and not only for the existence of it, we must

    Thus, some day in the future we will (hopefully) have an HTML5 test suite, consisting of a few hundred thousand sub tests. That would not be possible unless there is versioning.

    Or WaSP (or someone else ) could build a test suite that falls in between of a full spec test and the rather pointless “HTML5″ support charts that we see in abundance today. I.e, one can build a test that tests for a reasonable subset of the spec, and have that test be an indication about the quality of an implementation.

    Of course, this can be done with the WhatWG model as well, since it stipulates that some parts of their spec is mature and some parts are not. However, from a pedagogic point of view, that is an inferior approach.

    2. As has been pointed out by others, the W3C is also essential because of their patent policy. Thus if an entity, say Apple, gets a patent for a CSS-related technology – which they do on a regular basis since the US patent system is broken beyond absurdity – that at least makes that patent free for all to use, if Apple wants it to be a web standard.

    If only for this, there is no way the W3C can be left out of the process. And this also requires versioning. Patent holders will NEVER submit their patents to a “living spec”. They must know exactly what they are contributing to. And if a patent holder objects to something in a standard, refusing to let that patent be used, that also requires fixed versions.

    Versionless is how browser vendors de facto implement HTML. No single browser has implemented to complete HTML 40.1. And even while Mozilla are ironing out the last remaining ECMAScript 5 bugs, they are implementing stuff that will go into the 6th edition, like proxies. Thus, the fact that ECMAScript has fixed versions does not stop browser vendors from innovating.

    And we are thankful. Otherwise there would be no Ajax, except hacking the iframe. There would not be an innerHTML property, etc. (But, yes, innerHTML should really have been a method, not a property, in a perfect world…)

    aside – but this topic has come up in some comments above:

    The W3C patent policy takes the sting away from software patents. And actually I’d not mind H.264 to be a standard on the web, if the MPEG-LA consortium just agreed to license their patent portfolio under W3C terms.

    However, MPEG-LA has consistently refused to free up their patents, thus by virtue of their own actions they disqualify themselves from providing a web standard technology. If you’d like H.264 to be part of the open web platform, please pressure your government to drop software patents from the law or pressure MPEG-LA to allow their patents to be used under W3C terms.


    3. Perhaps also a side note, but it is being used as an illustration in some comments. Chrome is not “versionless”. Web developers manage their own machines and they get the auto updates. However, Chrome will soon make its way into corporate environments. There already are pre-made MSI-packages and AD group policies to use. But no sysadmin in his or her right mind will allow for auto updating software on the domain they manage. Thus, auto updates will NOT happen. And sysadmins will NOT want to push updates every 6 weeks.

    Thus, when Chrome goes into the corporate world, we will see that same issue as we have with IE today. Users who can not use the latest version and who may not even know what version they are using. But in many cases it will not be the latest.

  57. Jeffrey,

    And HTML 1.0 was developed primarily be one person, too: Tim Berners-Lee.

    There’s no surprise that initial specifications are developed by individuals or small groups of people. When a specification has led to widespread implementation and adoption, though, all stakeholders need to have a say in how the specification evolves.

    Beyond version 1.0 the concept is no longer one person’s vision.

    There may be only a few listed editors for any of the documents, but the concept is that the editors reflect what the group arrives at by consensus, rather than the working group form nothing more than backup singers for a single rock star.

    The existing HTML5 editor is incredibly intolerant of disagreement, disdainful of anyone other than a small group of browser developers, dismissive of accessibility concerns, and openly derogatory of the W3C. By any stretch of the imagination, that the W3C continues to condone this shows it has failed its fundamental task as HTML custodian.

    Peter Winnberg earlier pointed out just one of the bad designs being enshrined in HTML5: the redefinition of elements, such as small. This element had one semantic meaning earlier, and now it has a different semantics. Or the existence of the s element–a hopelessly flawed decision, permanently embedded into the new über HTML Continuum.

    Yes, the W3C dropped the ball for years by focusing too much on XHTML. At the same time, the W3C was undercut during the same time by the browser companies, who worked to ensure that XHTML would not be a success: Firefox and it’s absurd XHTML error handling, and Microsoft’s lack of support for XHTML until XHTML is on life support comes to mind.

    Now the browser companies have instigated a fine mess reminiscent of the good old Netscape/IE early days when the only want you can ensure people can view your page is either stick with older specifications, or tell people, “This site is best viewed in…”

    It is to their advantage to have HTML in this shape: they can then focus primarily on competition with each other, without having to worry about cross-browser issues, or ACID tests. So what if the majority of web designers and developers are caught up in a never ending cycle of not knowing what works with what–some script kiddie somewhere is happy because, wow! Look at this cool new thing!

    Though I typically agree with Steve Faulkner, I disagree with him that the W3C is in control. It is not in control of HTML. It basically rubberstamps what the editor wants. The only way to get the editor to back down is to get one the browser maker reps, typically Maciej from Apple, to get him to back off.

    Do we really want HTML controlled by one man, and owned by Apple, Google, Mozilla, Opera, and Microsoft? What about those others dependent on HTML, such as those in the accessibility community? The tools development community? Web authors and designers? Creators of eBooks and eBook standards?

    And now, this “living standard” just muddies the waters even more–it is nothing more than an open challenge to the W3C. It says to the W3C: you may think you control HTML, but you don’t. All we’re giving you, is a snapshot, the illusion of control.

    And the rest of us just nod, and go “Cool!”

  58. One last note:

    Previous versions of HTML started with the existing version of HTML as baseline and then appointed editors.

    The new WhatWG approach assumes that the same man will be editing HTML until he no longer wants to continue.

    What about this isn’t giving control to one person?

  59. And the rest of us just nod, and go “Cool!”

    I was initially so concerned about HTML5 that I convened an in-person meeting of minds. The standardistas listed here flew to New York at their own expense to meet with me in my studio for two days, where we ignored our clients, publishers, and other work to hash out our concerns about HTML5, which we published, and which the WHAT WG (or Hixie if you prefer) subsequently addressed. Prior to that meeting and publication, we all studied up on HTML5, many of us having earlier blogged our concerns—and in my case also published articles such as John Allsopp’s Semantics in HTML5, which is hardly a love letter to HTML5 as it was shaping up at the time. I should mention that John was also a prime motive force in convening that meeting, which he was then unable to attend because his wife’s pregnancy rightfully took precedence.

    I’d scarcely call any of that just nodding, and going “Cool!”

    Nor were my friends and I exceptional. The path to acceptance of HTML5 has been tough for many people. Whether we come down on the side of acceptance, as I have, or in the “against” column, as you have, I hope we respect each other enough to recognize that much thought, pain, and grappling goes into either position. Those who oppose HTML5 aren’t fashion followers saying “cool;” neither are those like me who support it. Let us respect each other.

  60. Aaron G, you were right about the errant link in my post—fixed, now, thanks! I also agree with your opinion about milestones being needed (a great job for the W3C) even if Hixie and his merry band update “HTML” as an ongoing stateless process. I think of the milestones as snapshots of a work in process.

  61. The English-language analogy is apt. How far I can stretch it?

    While style manuals may prescribe “proper” English for certain formal contexts, modern dictionaries do not—they document the way the language is used in real life (viz. “ain’t”). HTML5 development has incorporated this model, documenting what works in real browsers, then basing recommendations on that.

    Dictionary’s milestones were based on the constraints of print publication, but that is going away. The Oxford English Dictionary now updates their online 3rd edition regularly, and might never produce another print edition.

  62. Jeffrey, you made a good point and I shouldn’t let my frustration goad me to the point that I deride those who have a genuine interest in seeing HTML move forward.

    At the same time, thought, I just don’t believe that we’re directly confronting the significant disconnects that can’t help but form land mines when it comes to HTML in the future. Among these, and most prominent, is the concept of “living standard” as compared to “snapshot”, the potential problems this duality can generate, and indirectly, what it says about HTML ownership.

    Even in your statement about the standardistas meeting in New York and issuing the Super Friends statement: it wasn’t up to one person to address your concerns, was it? And were they really addressed?

  63. @shelleypowers I understand the perception is that the whatwg via hixie control the spec and thats one of the reasons it’s all hats off to hixie, people including standardistas love a perceived winner.

    There is still a way to go yet in the development of HTML5 as a standard and that is happening at the W3C. Unlike Hixie’s faux ‘last call’ last year on the then whatwg html5, the W3C Last call expected to occur in May will result in scrutiny and comment from a wide range of people and organisations. The html5 editor is obliged to make any changes to HTML5 decided by the HTML WG and that will continue during last call, unless the whatwg/he wants to create a fork of the specification.

    In regards to Ian as the editor of WHATWG HTML for life, thats for the browser vendors (except Microsoft of course) to decide. What we can be fairly certain of is that he will be continue to be the editor of W3C HTML5 standard during last call unless his hubris gets the better of him. Whether he is the editor or sole editor of HTML5x/HTMLx is another matter.

  64. Shelley,

    I don’t see that having one person in charge is necessarily bad.

    I’d take it a step further — most Open Source projects have benefited from strong “Benevolent Dictators For Life”. Linux, Python, Django, Rails, the list goes on — wouldn’t be where they are today if not for their leadership being a) there for life (If you don’t like it, fork it) and b) having strong opinions.

    PHP is a faltering mess with varying degrees of kitchen-sink-ism because Rasmus Lerdorf isn’t really enforcing any kind of leadership. Perl is kind of at a dead end (generalization). All in all, “BDFL”s tend to be more good than bad.

    Assuming “single leader bad” is just that — an assumption. Assumptions bad. That’s the one assumption that’s valid here.

  65. Robb, Tim Berners-Lee made the decision early on that he wasn’t going to be the sole dictator of what happens with HTML. He made the decision, and contrary to current opinion, a good one, that HTML would not be under the control and direction of one specific person.

  66. Steve, you have more faith in the W3C leadership than I do at this point. I do not see things going either openly or well during the LC comments. I fervently hope I’m wrong.

  67. Shelley, I agree with you to some extent. For instance, the patent encumbrance or lack thereof on HTML5 is unclear because HTML5 is developing outside the W3C. The W3C has a clear, overarching “no patent encumbrance” policy on all its standards. The WHAT WG has no such policy. Neither Hixie nor the contributing companies (Apple, Google) have told the rest of us what is safe or unsafe for us to do with HTML5. At the whim of any company contributing to HTML5, we might be very unpleasantly surprised in the future. I don’t expect such surprises, but they are possible and will remain so until this matter is addressed. Clearly, it must be.

  68. Shelley,

    Just because TBL decided not to be a dictator over HTML back when it was first created doesn’t mean that Benevolent Dictators altogether are a bad idea, or even that having a BD for HTML now, 17 years later, is a bad idea.

    As another example of a BDfL I would suggest Apple; Steve is very much a Benevolent Dictator when it comes to running Apple, but in doing so for the past 12 years he’s created a culture where no small amount of magic happens. Whether you personally like Apple’s products or disagree with them on some ideological or fundamental level, the truth is that Apple’s products typically have the highest consumer satisfaction ratings worldwide, consistently. It’s just about the penultimate example of how Benevolent Dictatorship can be much better than design-by-committee.

    As for the living standard aspect, I propose looking at it in a more organic, evolutionary perspective: is a website ever truly “finished?” Are websites themselves better left alone, rather than constantly nurtured? I say no.

    Very few websites get made in order to never be touched again; HTML as living standard simply reflects that principle in a very core, foundational kind of way.

  69. @Peter Winnberg, @Shelley:

    Because it redefines the semantics of existing elements like small to something that is semantically incompatible with the semantics it had in previous HTML versions

    This “elements being redefined” argument doesn’t stand up. The small element has not had its semantic definition changed, because it never had one in the first place. Its purpose was “small text size”, which is presentational, in no way representing the nature of the content, so it was semantically worthless.

    As I understand it, small has now been given a semantic meaning based on its typical usage (small print, e.g. copyright warnings). Similarly, b now means “stylistically offset text”, because it’s always been used to stylise text bold. Other things like big to which this principle couldn’t be applied were declared obsolete. You can agree or disagree with this, but either way it’s not the same as taking an existing semantic element like ul and totally changing its meaning.

  70. David Goss:

    You are right about one thing. In the HTML community people usually say that presentational only elements are semantically worthless. However they still “mean” something, apply a certain style to the selected text.

    But let’s try to explain what I mean using an example. I have a collection of HTML 3.2 documents that I want to upgrade to something newer by changing the doctype. One of these files contains this markup:

    <h1>About my love for <small>small</small> type</h1>

    The reason why I used the small element here was because I thought that it would be quite funny if the small word was smaller than the rest of the heading.

    Now if I change the doctype to HTML 4.01 then it will have the same meaning (heading level 1 where one word is smaller than the rest) and the validator will not say anything about that part ( could complain about other parts of the documents of course ).

    However if I change the doctype to HTML5 it will not have the same meaning anymore, however it might look the same. It will mean something I didn’t intend it to mean. And a validator cannot help me find this problem because it does not know if I used the small element correctly or not so it will gladly accept it as valid HTML5.

    Now if HTML5 instead had deprecated the small element because it is presentation only and added a new element called sidecomment like I suggested earlier a validator could recommend two options when it finds that line of markup. Recommend that CSS is used for changing the font size instead of the small element. But also inform the user about a new element for side comments which could have been one, of many, uses for the small element in the past.

    If you want to be compatible with other versions of the markup language, then the meaning of an element can only be changed when the new meaning includes 100% of all uses of the old meaning. This applies to all elements; elements that are presentation only are not excluded from this.

    This all happened to be the small element, there are also other problematic changes in HTML5.

    But I guess another way of fixing this would be to say that HTML5 isn’t compatible with other versions of HTML and redefine things exactly as we want it. But wasn’t that exactly what XHTML2 tried to do?

  71. We all just want Open Standards. And everyone is just going to argue about the semantics about how we refer to it? Really? This is retarded. How about this: To the oblivious average human, and to the ignorant CEO, it will be called HTML5. And amongst web designers and people in the know it will be referred to by what it technically is: HTML, CSS3, etc.

    Didn’t A Book Apart publish HTML5 For Web Designers??

Comments are closed.