Building upon untested assumptions

This morning printed an “Interview with Ian Hickson, editor of the HTML 5 specification”. A particularly disturbing section:

Bruce Lawson: You’ve said that HTML 5 is in “direct competition with other technologies intended for applications deployed over the Web, in particular Flash and Silverlight”. Why is it so important to do so, and isn’t it a lost cause given that those techologies are already out there while HTML 5 is not yet complete?

Ian Hickson: HTML 4 is also in direct competition with proprietary technologies, and it’s winning, hands-down. HTML5 is just continuing the battle, because if we don’t keep up, then the proprietary technologies will gain ground.

Bruce Lawson: What are the main philosophies of HTML 5?

Ian Hickson: Backwards-compatibility, incremental baby steps, defining error handling. Those are the main philosophies.

All day long I’ve been trying to puzzle out how someone could fit those sentences together, what type of ideas must first be assumed in order to make such an utterance conceivable, much less worth saying in an interview.

Tweezing apart some of the apparent assumptions of anyone able to make such a statement:

  • “HTML 4”, as a cross-company hypertext specification, is said to be somehow directly comparable to a common clientside capability such as Flash. How can the HTML 4 spec compete with its own extension mechanism? What must you assume before being able to craft such a sentence?
  • Aside from issues of logical levels and comparing apples to oranges, there is the apparent assumption that “anything Flash can do, my favored runtimes must do, better”. Why? If Flash has added many capabilities to today’s World Wide Web, why must people billing themselves as “The Open Web” fight it? He seems to think there’s obvious reason, but there’s a step missing in the description.
  • Perhaps most importantly of all, he’s assuming that a hypertext specification should become a multimedia/RIA specification. This is dangerous to the entire HTML ecology. Even the current specs are confusing, after they’ve spun out styling, object model, even image formats into separate specs. Stuffing an additional drawing spec and a storage spec and animation and 3D and what-all into the hypertext spec weighs it down even further. I think he is espousing a recipe for ruin, and I don’t know why, because he doesn’t publicly examine his assumptions.
  • The mention of Silverlight is intriguing. This cross-browser plugin has had no realworld effect. Silverlight does not realistically threaten Chrome. HTML5 and Silverlight perhaps compete in blogosphere mindshare. Or perhaps he’s just raising it as an “M$ boogeyman” to better bash Flash. There’s something that prompts him to include Silverlight in his assertion, but it is not clear what assumptions underlay its mention.
  • Why even bother to say “and it’s winning, hands-down”? Is there a “winning”? If so, how do you measure it? What type of boosterism could produce such an utterance?
  • Bruce directly asked whether this was “a lost cause” because Flash was already in use while HTML5 was still in committee. Ian ignored this vital question of how to bring a concept out into the world, making a capability genuinely useful to people. It seems he assumes it’s not important enough to think about.
  • He assumes that “HTML5” is “continuing a battle”. Perhaps he knows what the battle is. The rest of us have to guess at his assumptions.
  • He assumes that there is something called “proprietary technologies”. I assume he does not assume this includes Apple’s Webkit governance, or Google’s ubiquitous web beacons.
  • The “main philosophies of HTML 5” include a helpful point: defining what browsers “should” do when they encounter improper content. But “incremental baby steps”!? It’s more like a wild jump to Rich Internet Applications. Maybe the assumption here is “No one will challenge me on this jive”.

I’ve been trying to see the world through Ian’s eyes, but I cannot — he has built a rhetorical structure atop untested assumptions, instead of establishing upon common ground.

What would I like to see?

I’d like to see some clear discussion on the appropriate scope of a spec that any hypertext implementation should be able to easily reach. If CSS is a separate specification, then why isn’t RIA?

I want to see more “standards” discussions about the needs of people who actually publish to the web. We should not have had the last ten years of people working around browser differences. No more!

I wouldn’t mind seeing standalone RIA specs in the SMIL, HTML+TIME, SVG mode. These live or die on their own, and do not threaten the extraordinarily-useful HTML spec itself.

I want to see “open web” evangelists find ways to bring their functionality into other browser brands. Instead of worrying about microshare, figure a way to make cheap codecs available to any person in any browser today. Browsers were opened up to third-party rendering long ago — don’t fight to turn back the clock. Stop railing against cross-browser functionality; stop thinking you must own it all yourself; if you say you’re for “open video” then act it. A campaign for a Theora plugin would hold more relevance.

(The proposed fragmentation of Today’s Web doesn’t worry me as much… Flash makes too much sense for people to abandon it, and Adobe thrives by uniting fragmented silos. Those content developers who don’t see through the hype will be the ones to pay the price. But increasing fragmentation and siloing would not be a good thing for HTML in general, which is why I’m writing this essay.)

What do I think is really going on?

I think it’s a corporate powerplay to “bless as standard” runtimes from the corporations pushing the multimedia syntax. There’s no angle for realworld developers in VIDEO tag and the rest — their proponents shy away from any discussion of how these might someday be supported and useful in the world. This campaign is not for content developers or their audiences. It’s a marketing play, like how Microsoft lobbied for OOXML.

That’s why it doesn’t matter that VIDEO leaves codecs unaddressed… why it doesn’t really matter that most consumers use Internet Explorer. This is not really about content developers. It’s about browser vendors. They can make any content requirements they wish upon publishers, once their engine is “blessed” for VIDEO tag.

Some vendors sell proprietary hardware. Some sell proprietary data about what you watch to advertisers. I suspect either would be happy to fragment Flash capabilities if they could. Instead they’ve found other ways to construct a walled garden around audiences. And it’s easy enough to find fanboys as footsoldiers.

Flash scares them because it opens things up too much, levels the playing field. That’s what I think is really going on.

Summary: The HTML5 editor says he’s fighting a battle against Flash. But he doesn’t explain why, so it’s hard for us to help him get better.

[Comments: I’m not really keen on diluting this conversation with guesses about what he may have meant, thanks in advance.]

9 Responses to Building upon untested assumptions

  1. Adrian says:

    I think you’ve over-reacted, majorly. [jd sez: I don’t know who “adriansuch@hotmail” is.]
    Vendors and Developers should focus on the best technology for the content, be that HTML, CSS, Flash, SQL, XSLT, XVG, WordPress or (dare I say it) Silverlight. In an ideal world, you could write/code once and be done and everything would be open-source and free.
    Instead the world has Dick Cheney and more web “standards” and versions than minutes in Police Academy films.
    There is nothing wrong with some healthy competition eg Firefox, Flock, Chrome, Safari, Opera in the web browser space. The same can be said with major web technologies.

  2. Desmond Hume says:

    JD, you have some valid points. I think HTML5 is reaching very aggressively into the RIA space. I had thought that the WebApp/WebAPI initiatives were better aligned for this, and I feel your “what you would like to see” points 1-3 are particularly valid. I agree that there is a corporate power play at work here, to a certain extent..
    However, you must understand the sentiment of the open source/technologies effort. Despite the history of Flash and Flex as Macromedia products, you’re now part of Adobe, an organisation that, for justifiable reasons or not, behaved in an exceedingly aggro manner. The standards and open source folks aren’t just going to forget Adobe’s rather atavistic pre-merger behaviour: “PDF is the platform”; Adobe’s attempts to support and promote only standards they owned – all anti-standards.
    That is a rep that sadly spills over to the Macromedia-developed product line, but one you nevertheless have yet work to do to overcome.
    One should note that the aforementioned corporate power play is being fomented by an organisation much large than Adobe, yet using the same 1980’s desktop software ‘playbook’, and it is unfair that this organisation’s machinations are so conveniently ignored. Post-merger Adobe should not have to fight a wretched pre-merger image whilst the biggest player is only acting in a way that makes Adobe’s behaviour mild in comparison. Yet it is what it is.

  3. Isofarro says:

    JD sez: “This is not really about content developers. It’s about browser vendors.”
    Considering one of the major focuses of HTML5 is to accurately or appropriately document error-handling and DOM implementation to ensure stable behaviour within browsers, you’d be quite accurate with this statement.
    I find your suggestion of a Rich Internet Application working group curious considering the fractious history of what is now called the HTML5 Working Group. (See for example Sam Ruby’s HTML 5 Evolution: )
    I believe the key point in time that’s directly responsible for HTML 5 and it’s anti-proprietary technologies is that meeting of the W3C Web Applications Activity to discuss a baseline set of technologies to form a Rich Internet Applications platform way back in 2004.
    At that particular meeting larger organisations: IBM, Microsoft (can’t remember what Adobe’s proposal was, it might have been PDF) essentially declared that the browser was dead, and working to expand HTML to meet the needs of web applications was not viable and should be done outside of the W3C under a different name.
    At that point individuals from the two browser vendors (Opera and Mozilla) – mainly from Ian Hickson (who was working for Opera at that time), supported on the sidelines by Apple decided to do just that. That’s where WHATWG formed: Web Hypertext Working Group. Their specifications included WebForms – basically X-forms that could be used in HTML4.
    Over time the scope of their work increased. And somewhere along the line the W3C invited WHATWG to continue work within the W3C.
    I don’t recall anything relevant coming out of the Web Applications group in the meantime. So it suggests that they’ve made no progress since the browser vendors essentially walked out.
    Microsoft. Well, that’s always been the stumbling block no matter which direction things went. When the HTML5 working group formed, then Microsoft threw their hat into the ring with Chris Wilson originally chairing the HTML5 Working Group.
    Just off the top of my head. Interesting piece you’ve written.

  4. Ed Everett says:

    ‘”HTML 4″, as a cross-company hypertext specification, is said to be somehow directly comparable to a common clientside capability such as Flash. How can the HTML 4 spec compete with its own extension mechanism? What must you assume before being able to craft such a sentence?’
    As a web designer and developer I need to decide which technology I should use to create each part of the website. Often the choice is between Flash and HTML/CSS/Javascript – as such flash is in competition with HTML. The assumptions you need to get here are that the developer is thinking about the best way to deliver content to the website’s audience.
    A situation where they are directly comparable would be, for instance, a slide show. There are many Flash solutions available and many HTML/Javascript solutions. I compare them and choose one that fits my needs.
    This competition is healthy. To create a decent multi-file upload experience at the moment needs Flash (or some other plugin), but Browsers have seen the need for this and some have started to implement native solutions. This is good for the end users as a good user experience becomes easier to implement.

  5. Brad Neuberg says:

    [jd sez: My apologies to Brad… this got flagged as spam by the filters, and I’m 11 days late in catching it and sending it through! 🙁 ]
    IMHO, the process with respect to Flash and HTML 5 is exactly how things should work. Folks like Flash, Gears, JQuery, etc and others come up with new innovations, and these slowly trickle down into standards that are then implemented as a common base. Things like Flash then innovate with a new round, such as the PixelBender technology. This is a good dynamic that aids both innovation and standardization, a kind of pendulum that swings back and forth.
    Web video is a great example of this; RealVideo and Quicktime developed the first iteration, Flash revolutionized it by making it easy for viewers to use, and now, in 2009, web video is essentially a commodity and should become a base part of the platform. It’s time for Flash to move onto the next step, which it already has been doing with things like Air, RTMP, etc.
    Also, about RIA and HTML, in case no one noticed… JavaScript and HTML have _long_ been subverted away from their original text and hypertext-only intention (starting in about 1997, or perhaps even when the IMG tag was baked in) and have been an RIA technology for a long time of gradually increasing capability.
    Iterating existing systems like HTML towards greater needs has been _exactly_ the (successful) history of the web: . Tim Berners Lee invented the brilliant (simple) evolvable trifecta of HTTP, HTML, and the URL and then promptly forgot all those lessons with the founding of the W3C…
    I’d love if Flash become an Open Web technology, and have written about this before: . Flash is still a bit of a black box, even with ExternalInterface, and needs to come a bit closer to the web’s way of doing things (basically, less doing everything at compile time ala Flex and more doing things at runtime). To be honest, the ball is in Adobe’s court on this one. I’d love to see a response to that blog post I wrote awhile back. Yes, Flex is open source (thanks BTW for that), but thats too high up in the platform layer cake to really matter.
    And yes, I agree with cross-browser rapid extensibility. That’s a _big_ hole in the web today that Flash admirably (and very successfully) fills.
    Brad Neuberg

  6. Cam says:

    The reason so many in the “open source” world are upset about the proprietary standards is because they ultimately leave someone in the dust. Right now, I’m running Ubuntu, 64 bit, and the one thing that always pisses me off about the flash plugin is it’s insistence on being a total pain in my rear. I’ve still not figured out why Adobe likes making sure some of their ‘standards’ don’t work for vast swaths of internet users. As a web developer myself, I don’t even get to choose whether to use Flash, Silverlight, or HTML/CSS/JS, I only get the option of using HTML/CSS/JS 🙂
    [jd sez: We readers can’t tell what your problem is… I’m not even sure if you’re using the Ubuntu-happy 64-bit Player.]
    I can’t wait for HTML5 to arrive. It has the backing of most of the browsers, so if they can keep the heat on Adobe and Microsoft, even better.
    [jd sez: Apple, Google, Opera and Mozilla all together support only a minor fraction of the people that are supported by Microsoft browsers. Instead of asking “When will HTML5 arrive?”, try thinking about which levels of which HTML5 will be supported by which audiences… more realistic.]

  7. Somewhere you are right but I feel you have over-reacted.

  8. Bill Reed says:

    Let’s rename “HTML 5” to “Web Browser 5”, because the open process that HTML 5 claims to be is in fact a charade. All the influence and decision making power is help by the 5 Web browser vendors (Mozilla, Opera, Apple, Microsoft and Google). The fact that screen reader vendors do not want to participate in the HTML 5 process is an indication of how marginalized their opinions will be. And others who are passionate critics of the HTML 5 process get punished through bulling on the mailing list or banned from the list.
    And is it a mere coincidence that Hixie started to author the HTML 5 spec as a Google employee while Google was secretly developing a Web browser?
    In my opinion, the only reason Hixie can say, with a straight face, that HTML 5 will not perpetuate bad markup, is because he legitimized bad markup in the HTML 5 spec. What used to be “bad” is now “acceptable”.

  9. Kit Grose says:

    The problems with treating Flash as the go-to method for certain technologies are clear to me:
    1. We’re all at the mercy of Adobe’s developers when it comes to memory usage, performance, etc. While this seems like needless idealism, anyone who uses a Mac knows Flash isn’t exactly a speed demon cross-platform. It also directly affects the web as it moves toward mobile platforms. The gorilla in the room here is the iPhone, for instance. Also, I think it’s fair to say Flash causes more crashes (at least in WebKit-based applications) than anything written in pure JavaScript.
    [jd sez: I use Flashblock in Firefox, and Safari for text sites. Both still beachball with regularity.]
    2. Linked to the above, competition is vital to the web space improving. Content authors are eagerly awaiting the features in HTML5 (though we know we’re looking at around 5 years before we can reasonably expect their use), and new features may (we hope) drive adoption of new browsers. It’ll also keep Microsoft releasing versions of IE (I’m of the school that feels the long delay between IE6 and 7 caused users to neglect to update to IE7 and 8 more than anything else). Browser improvement benefits us all more than the changes in HTML5 alone. Look at the incredible improvements in JS performance in V8, JSCore and TraceMonkey lately to see where standardising the spec could lead us.
    [jd sez: Competition’s great. Build a plugin, reach all browsers. Don’t perv HTML with RIA dreams.]
    3. Flash authoring is limited to Mac and Windows, and the authoring tool isn’t perfect. As a Mac user I’m not affected by the first part, but I can tell you right now, I hate opening Flash on my machine literally every time I open it. It might be great for some people, but not for me (I must admit, I find the pre-Macromedia Adobe apps orders of magnitude easier to use). When the outcome isn’t tied to a single authoring environment, content authors can choose to use whatever development environment they choose.
    4. System-specific implementation is a big one; if I build my calendar control in Flash, it’ll look the same on a mobile as it will on a 30-inch desktop PC. That’s part of the point. But when you define the control in terms of what you’re trying to accomplish, browser vendors can make more informed decisions based on the platform the content is being viewed on (like a platform-native calendar control). While it could be argued that Adobe could make themselves responsible for performing this task, it’s unrealistic to presume they will (who has the stronger vested interest in a positive user experience in, for example, the iPhone browser; Apple or Adobe?).
    5. As features are added to the core featureset of shipping browsers, Javascript frameworks we already depend on like Dojo, YUI and jQuery can begin to provide platform-native controls when the browser supports them and fall back on their custom implementations only when necessary.
    Basically, it appears to me that you’re suggesting Flash become a defacto standard, but (largely for the reasons above) that can’t result in a net benefit for users.
    [jd sez: uh, Flash is already quite the de facto standard. That’s what attracts publishers… it works for their audiences.]