Comic Update: HTML5 Manners

May 04, 2009

I’m going to lay out a chronology of prior events for you all so that today’s comic has a context other than the poor movie experience that was X-Men Origins: Wolverine (I really wanted to love that movie.)

Chris Wilson (W3C HTML WG co-chair and Microsoft employee) posted an e-mail to a HTML5 discussion that made reference to the “W3C HTML5 Spec”.

Mark Pilgrim (Google employee and WHATWG Blog author) in the WHATWG IRC channel then implies that HTML WG co-chair Sam Ruby would have been attempting to be divisive had he written that e-mail, but since it came from the other chair, Chris, he was in fact being stupid.

Shelley Powers (computer book author, software developer and technology architect) expresses utter frustration in a blog post about the future of HTML5 by pointing out this incident and many others that indicates a “Hatfield-McCoy feud” (in her words) between the W3C and WhatWG that is miring the whole process down. Gems in her post include an IRC discussion (starts here, ends here) between HTML5 editor Ian Hickson and Microformats champion Tantek Celik where Ian shows his bias in the microdata issue (read that: whether to include RDFa in HTML5) by asking Tantek to vet the use-case submissions. The “vetting” quickly devolves to the pair saying “Use microformats for everything” or if such a situation isn’t possible, to simply create a custom microformat for your own use.

Yes, that’s it, let’s make dozens of one-shot formats to solve the many microdata issues we’ll doubtlessly be facing in the next several years. That can’t possibly create any sort of data-harvesting compatibility issues. If I can see the shortsightedness of this issue (and I fail to wear coats on cloudy days because “it’s not raining yet”) then you can bet this isn’t a tenable, long-term solution.

They take some time to attack Creative Commons while they’re at it.

These aren’t the only times these sort of offensive public conversations have occurred, where WhatWG members have publicly derided, insulted or challenged the intelligence of the individuals they’re politely talking to in other conversations about topics they’re mutually involved in (such as HTML5). Mr. Last Week in HTML5 is a great (albeit foul-mouthed and somewhat spiteful) source of links to these conversations occurring all the time.

Ian responded to Shelley’s post, taking umbrage (as Shelley put it) at her “insulting accusation”. Shelley’s response cut to the core of the matter, exposing the main issue at hand here, and one that needs some serious addressing. In her words: “Don’t you get it? Don’t you see what Last Week in HTML5 is trying to demonstrate? You talk respect in my comments, or Sam’s comments, and elsewhere, but you show disrespect to me, to Sam, to others, in the IRC, and it completely undermines everything that you do.

I can’t state it better. These people aren’t average developers trading insults about trivial code snippets on small-scale projects. These are industry movers-and-shakers who are supposed to be working together to help create the standards that will define how we use HTML and other web technologies for years to come. I expect professional disagreement to occur (I’d be worried and concerned if that didn’t happen). But to start insulting one another personally in a public discussion (or frankly, privately) is shameful to the entire process and the entire community that is depending on them to do a good job.

Shame on you, sirs.

I’ll leave you with the following quotes from this IRC discussion including Doug Schepers, Ian Hickson, and a person named ‘roc’ (I don’t know his real name) [edit: As I've been informed in the comments, roc is Mozilla's Robert O'Callahan]:

shepazutoo (Doug): wow, Hixie, “contradicting other specs has never stopped the SVGWG before” (q.v. xlink, css, etc)… first, those were almost certainly mistakes rather than purposeful contradictions, and second, you’re acting like the current SVG WG is the same set of companies and individuals that wrote the SVG 1.1 spec, which you know to be false… can you please drop the political histrionics? we’re acting in good faith to correct some past errors, and to work with other WGs and with browser vendors to make all the specs align usefully

Hixie (Ian): i think you may have missed the smiley

roc (Robert): a smiley is not a “get out of jail free card” to be annoying

Respond To This Post

Share This Post With Others: |

Category: Comic, Drama | Tags: , , , , , , , , ,

33 Responses to “Comic Update: HTML5 Manners”

  1. roc = Robert O’Callahan

  2. roc is probably Robert O’Callahan of Mozilla.

  3. @Jeff & Philip – Thank you very much for the name!

  4. “WHATWG Blog author” is a bit misleading – anyone can submit an article to the blog, and anything non-spam should get posted. Mark Pilgrim has no unique status on the blog; he’s just a person who’s written a series of useful articles on there.

    It’s probably not helpful to view him as representative of the people working on HTML5 – he hasn’t directly contributed much to the specification (he’s not even listed in the acknowledgements). And he often says things on his blog/Twitter/etc that people find offensive, amusing, accurate, or some combination thereof, so it’s unsurprising when he does the same on IRC. The people in #whatwg don’t find it unacceptable behaviour because they know that’s his style and don’t take it too seriously, but it looks more like a straightforward insult when taken out of context.

    (In this particular case I don’t understand why he objects to what Chris Wilson said, and he failed to give any justification, so I viewed his comment as merely useless and unhelpful (like much of the conversations on IRC – that’s just a consequence of the medium) and easily ignored.)

    In the microdata discussion, I see Hixie requesting feedback from someone with relevant experience in the issue of marking up data, expressing concern that the use cases put forward were focused on justifying RDFa rather than on solving the problem in the best way, expressing a failure to understand to why people choose a complex solution to a problem rather than a simple solution that seems to work just as well for them, expressing frustration with people who are promoting those complex solutions, and expressing personal opinions on licence proliferation. I don’t see anywhere he “publicly derided, insulted or challenged the intelligence” of anyone.

    Maybe he’s too incompetent (or acting in bad faith) to come up with a decent solution despite all the time he’s spent discussing the issues and collecting feedback from as many sources as possible, but that would be a very different complaint and doesn’t seem to be what people are complaining about (particularly since he hasn’t yet proposed a solution to the microdata use cases). So I find it hard to understand what the cause of the apparent problem is or how it should be resolved…

  5. @Philip Taylor – I’ve seen Pilgrim referenced as “WHATWG Blog author” numerous times when double-checking his identity, so made the (incorrect) conclusion that he was largely responsible for its content. It’s my mistake for drawing too strong of a connection, and I apologize for overstating his role.

    With the microdata discussion, it was meant to illustrate an imbalanced level of dialogue/bias in regards to how Ian is handling Microformats vs RDFa (I’m not saying those formats are in competition with each other, although it’s true that many view them that way). At the very minimum, it seems to portray that he is acting in bad faith in regards to devising a solution for RDFa and HTML5 living together.

    In reviewing the post, I probably should have arranged for that paragraph to come after the one discussing offensive conversations (which was meant to refer to Mark’s comments and those like it), as the bit about microdata was meant to illustrate the feud that is occurring, and which Shelley referenced. In its current positioning it creates the wrong impression.

  6. What makes you think he’s acting to devise a solution for RDFa and HTML5 to live together? He’s explicitly *not* doing that, nor does it seem obvious that it is incumbent on him to do so. (Hence, the “bad faith” claim misses the mark.)

    As he has said many times in general and specifically on this issue, his methodology is to start with use cases and (application) requirements (so no, “we require RDFa”) then to explore the solution space. Part of, AFAICT, Ian’s frustration with RDFa advocates is that, by and large, they push RDFa per se.

    And if you think RDFa is a good solution to a slew of important use cases, it’s easy to see how one might conflate the two (i.e., pushing for solving certain use cases and pushing RDFa per se). But I also think it’s reasonable to be skeptical about accepting *any* technology esp. when people want it built in (or have hooks added) to such a widely used and complex standard as HTML.

    So, I think Shelley and you are just misinterpreting that IRC conversation. AFAICT, Ian has spent quite a lot of time soliciting input from RDFa advocates…more, perhaps, overall than from the microformats community.

    If one is going to make an accusation of problematic bias, then it really behooves one to do a careful job documenting 1) the bias and 2) the problem with it. (Of course, if the problem is *outcome*, that is, the only acceptable *process* is one with the *outcome* of RDFa being supported in HTML5, then that should be clear up front. It’s a very clear bias, after all.)

    (If one doesn’t, how is one any better than Pilgrim? At least it’s clear he’s being over the top. Accusations with the appearance but not the substance of sober reflection and care are far more likely to mislead people, don’t you think?)

  7. I won’t defend Mark’s remark, which I found unconstructive, but I’d like to comment on the appearance of relations towards microformats and RDFa:

    I see Shelley’s point that if there’s a mismatch between non-IRC communication and IRC remarks, the non-IRC communication is undermined. That mismatch may have a lot to do with perception than intent, though. (Of course, perception matters.)

    People in the HTML5 community see issues with microformats. (The previous sentence is probably vague enough to be a truism.) In fact, three XTech conferences in a row, #whatwg regulars (Anne and I) asked after microformat presentations when they are going to have proper specs (well-defined processing spec and well-defined authoring conformance). The reason why public communication with Tantek looks different from public communication with some RDFa advocates, including Shelley, is that Tantek’s use case-oriented approach fits the #whatwg culture better than the stance that RDFa should be “in” because RDF is presumptively good / useful / starting point / Recommendation.

  8. Henri, you have good points, though I’m not sure what is a “good” use case. I’ve seen several listed by the RDFa groups, but none seem to measure up. When we ask why they don’t measure up, we’re told that they don’t, period.

    Yet, I’ve seen Tantek basically reduce the discussion to open vocabularies, bad, closed vocabularies, good, but such a reduction isn’t facing the same challenge by Ian.

    It does not help when I see several references to how “bad” RDFa is, made by Ian, not Tantek. Obviously, RDFa will never have a home in valid HTML5. So, I think we’re heading into–disdainful and mocking joking about validation badges on the IRC aside–when valid markup is no longer a goal, or even a viable alternative.

    That is the way it is going to be.

  9. Frankly, what I find more disturbing is that Ian gathers the use cases, he vets then, he rewrites them in his own words, and documents them, he generates proposals, he determines what will get incorporated, he writes the spec…

    How on earth did one person get so much power in this spec?

  10. Shelley, a number of use cases from the RDFa community ended up on the use case list. It doesn’t follow that RDFa is the best way to address those use cases.

    As for statements about RDFa being “bad” (URL?), have you considered the possibility that RDFa might actually be a bad design for addressing the use cases? (In the previous sentence, I don’t intent to take a stance either way.)

  11. It doesn’t matter Henri.

    Right now I have RDFa incorporated into my sites. I am using just one of multiple libraries that can extract the RDFa out, and insert it into a permanent store, and another library that provides a querying mechanism. I’m using three vocabularies now, and by the time I’m finished with additional uses of RDFa, I’ll be using another five vocabularies.

    Not once did I have to negotiate with anyone. Not once did I have to do more than write some relatively trivial PHP to invoke libraries, and make minor modifications to my Drupal templates in order to accomplish what I want to accomplish.

    I have this capability now, not some year in the future when we finally realize that, wow, maybe this stuff isn’t so easy, after all.

    As long as what I have now is not broken, I don’t care what you all invent in the HTML5. Go for it — reinvent the world. Have a nice time.

  12. Shelley, you do realize that it doesn’t make sense to point out as a problem your doubt about having been listened to if you publicly indicate that the outcome doesn’t matter to you?

  13. @Bijan – It’s not required for Ian to include RDFa in HTML5. He is responsible, however, for being part of a fair process to determine the merits of including or excluding it from the spec. He provides little in the way of constructive feedback to the RDFa use cases, simply making (clearly overinflated) statements that 99% of the related emails weren’t use cases at all, and that the rest were useless.

    Hyperbole is great for a humorist, it’s not so useful from someone writing a spec. It’s my opinion (key word) that he entered the microdata issue with no intent of seriously considering RDFa inclusion. This would bother me less if he didn’t hold such universal control over the spec’s development. As it stands, a larger degree of constructive feedback and a conscious effort to not engage in lopsided public discussions with RDFa opponents would at the very least allow the public the opportunity to buy into the appearance of a fair process.

    Setting aside the RDFa/Microformats/HTML5 issue momentarily, my major point of the post was that the conversations that occur in the IRC channel, even when they involve non-central individuals like Mark Pilgrim, create an issue of perception with the rest of the web community. Ian and others claim to be working from a non-biased standpoint, and post respectful blog comments, then fail to censure insults or non-constructive comments in the public venue by Mark and others.

    Anyone on the wings is allowed to behave as much like a monkey as they please. But the people on stage, the people who are entrusted to devise this standard in a collaborative effort, need to be held to a higher standard. Failure to do so will result in a failure for the community to buy into HTML5 as a genuine spec and failure for the spec’s approval process to go along smoothly without objection when it’s time for it to reach Last Call.

  14. Henri, the outcome mattered to me at one time. It seemed as if you and the RDFa folks were seriously working on how RDFa could be included with minimum impact in HTML5. It seemed to me there was willingness to compromise in both groups. You even incorporated a test case for RDFa.

    Then, everything just stopped. I have never seen such cooperation between groups killed so quickly before in my life.

    Now, it’s obvious RDFa won’t be integrated into HTML5. Does this mean I will stop incorporating RDFa in my sites? No. It just means that I’ll either stay with XHTML 1.1, go with XHTML 2.0, or use RDFa with whatever bits of X/HTML 5 I find useful, and say to heck with validation. Frankly, the W3C doesn’t care about valid markup, why should I? As long as whatever I use works in my target environments.

    Now, all I care about is that there is nothing incorporated into the HTML5 specification that would cause HTML failures to occur if we continue to use RDFa along with the HTML5 bits that interest us. For this to happen, though, for RDFa to be deliberately sabotaged in HTML5, would require a deliberate act of pettiness.

    As for the earlier comment about misinterpreting the IRC, should I link to section where someone asked why I’m angry all the time? Or the section where my use of validation badge in my personal weblog post generated so much humor among the IRC participants?

  15. Shelley’s documented experience around the WG’s response (to RDFa) is very similar to the reception that the Accessibility Community received early on, and still remains to this day. It boils down to which ‘team’ you subscribe to, and as Shelley pointed out in her blog, there is a Hatfields and McCoys atmosphere that prevails. Often the WG will emerge with a ‘solution’ that will tolerate no further scrutiny, as it has been hatched, discussed, and decided upon in the IRC back-channel, a space generally populated and controlled by a small band of browser employees with their own agendas. Yet when other expert communities apply the same approach (internal discussion with a final ‘pronouncement’), the WHAT WG cabal deride the comment(s) and demand further ‘proof’. The prolonged fight over how to deal with @alt is a perfect case study there. Shelley’s other point regarding an inconstancy (I call it two-faced) approach in communication (mail lists versus IRC – the topic of Kyles comic) is both highly accurate and a real source of much of the frustration surrounding the current ‘process’.

    It should also come as no surprise that Ian and Tantek are buddy buddy – they both are listed as ‘editors’ of the CSS 2.1 spec, another document that had a certain level of acrimony attached to it’s creation – Ian does have this habit of leaving a trail of disgruntled folk in his wake. (Funny thing about synchronicity: – from today).

    The really sad thing is that as the W3C continues to tolerate this garbage and allows this process to continue in their name, they tarnish the W3C’s ‘good name’ and reputation, and leave many wondering if they should bother worrying about the W3C at all.

  16. For the record the 15,000 to 150 ratio was not hyperbole. I actually rounded down because I didn’t have the complete numbers. I ended up with about 200 total lines in my use cases file (which I then wrote up and e-mailed the list), about 50 lines of which were blank lines (separating use cases), about another 50 lines of which were things I’d come up with on my own. These use cases came from about 15,000 lines (actually more like 16,500 lines, but I forget the actual number — it was definitely more than 15,000 though) of e-mail from the last 6 months, plus all the comments on Shelley’s blog entry, plus all the stuff Manu wrote up (Manu is one of the few people to have actually done any real work to help here, and I am very thankful to him).

    So it really was about 15000:150. And that’s only after sending an e-mail saying “I am specifically looking for use cases only”, I’m not even talking about the threads before that.

  17. @Ian – I retract the hyperbole statement regarding your use-cases statement, then. My apologies. Although I’m going to agree with Shelly that harvesting comments from her blog to serve as a source of use-case requirements isn’t productive. I imagine, as that material isn’t intended as a formal submission, that on average it will only help provide chaff and not wheat.

  18. @Kyle. So, I’m confused where we stand. Or, perhaps, where you stand.

    It seems to me that a lot of your substantive complaints dissolved upon closer inspection (do you agree?), e.g., hyerbole. Now, perhaps that supports your argument, i.e., that there is a PR problem. It’s certainly the case that PR can matter to the success of a spec, so this is a potentially substantive point.

    But we can decide, a bit, the world we live it. Are you quite serious when you write:

    “”"Ian and others claim to be working from a non-biased standpoint, and post respectful blog comments, then fail to censure insults or non-constructive comments in the public venue by Mark and others.”"”

    Ian is supposed to censure comments like that? Really? Does he have to censure Mark’s posts on Mark’s blog as well? How does the censure take place? Does he have to send an email? Refuse to talk to Mark? Disregard 1 of Mark’s desired features? Stop participating in any channel where Mark appears?

    Can you be a bit more precise about what you want Ian to do here?

    And you retracted your claim about hyperbole (and, I’ll note, you wrote ” He provides little in the way of constructive feedback to the RDFa use cases, simply making (clearly overinflated) statements that 99% of the related emails weren’t use cases at all, and that the rest were useless.”…so, they weren’t so *clearly* overinflated. Where did your judgement that they were overinflated (inflated?) come from? Was it the result of bias? Is that bias a problem? What kind of problem?), but do you also retract your claim about actual bias on Ian’s part? (Instead of perception of bias?) I mean, I would be interested in a substantiated claim of 1) bias that was 2) inappropriate and 3) harmful to the overall development of HTML and 4) replaceable with something better.

    (I put all these conditions on because I’m generally unimpressed by general complaints of bias. Of course, everyone has technical tastes and often there is *no objective* rank between alternatives, and even more often there’s no reasonably determinable rank other than taste. I think it’s acceptable to delegate lots of decisions to a person…we always do in the end. For me to worry about bias I need to understand both the substantiated harms of that bias and there has to be a sane plan for replacing it with something else. If we just substitute one bias for another without any *other* improvement, then it was a waste.)

    Now, obviously, these shifts in bias can asymmetrically affect different communities or individuals who, thus, may care passionately about their lot (or be passionately attached to their lucky lot). And there does seem to be ways that unlucky folks can change the situation: Implement features in browsers, have runaway success with their technology so it becomes significant on the web, participate in the HTML WG and raise an issue that convinces the chairs to open the issue to a vote and the WG to vote their way, or convince Ian directly (among other tactics). Some of these are easier or harder or more or less frustrating (for some people) than others.

    Hmm. I’ve no idea where I’m going with this :) I guess it’s that if one finds one’s position being dismissed, perhaps it’s because you didn’t make a very strong case.

    As a final point (while the flu kicks back in :)), I’m not sure what’s *wrong* with harvesting comments from blog posts. Who cares if they aren’t intended as a formal submission? Shouldn’t we actually look and see if they are chaff?

    Ian does this continually, afaict (and, frankly, amazes me that he has the energy to do such wide sorting of input). Shouldn’t it be, I don’t know, a *good* thing that he sees and seeks out input that *didn’t* get formally submitted?

    It just seems weird to complain that he’s biased *and* that he seeks out as much information to take seriously as possible.

    But that might be the flu talkin’ Oink-cough!

  19. Sadly, Shelley’s blog actually had a higher signal:noise ratio and a higher total number of use cases than the actual mailing list posts. I always take into account any use cases I know of, though, I’m not going to ignore use cases just because they were posted somewhere unusual. This seems like a good thing to me.

  20. Bijan, Kyle’s opinions are not discredited, just because Ian comes in and say, “uh uh, that’s not the way it is.” Since Ian has decided not to formally document the so-called “requirements” and “use cases” from which he supposedly derived his formally submitted use cases, we have only his word that he’s winnowed through thousands of words in order to present a pristine collection of truly fine use cases.

    Ian, actually quite a few use cases were submitted formally, but you kept replying that they weren’t really use cases, as defined by some criteria you set. We then asked, what is the criteria by which you judge a “good” use case, and you answered there is no criteria — you basically know what a good one is when you see one.

    In other words, everything you have done in regards to microdata is based on your biases. And since, you have in the IRC specifically mentioned your disdain for RDF and RDFa, that there was no fairness or objectivity in your effort is a given.

    You set the criteria of what makes a good use case, you then selectively go about gathering what you decide should be considered use cases, you then re-word them in your own words, setting them in such a way that they become nothing more than straw men, easy for you to knock down.

    I know you have never worked outside of standards work with browsers and Google, so you’ve never worked on applications in use at companies such as Boeing, Intel, and so on, but the requirements gathering process is handled in such a way as to ensure no doubts about the fairness, and the comprehensiveness of the effort. The end result of the effort isn’t to create a requirements document that addresses the document writer’s interest, but to reflect the needs of the community the application has to serve.

    All you have done, by managing all of this on your own, with no outside intervention, or even documentation from which you’ve supposedly derived your use cases, is generate an atmosphere of doubt and disbelief.

    I would say your effort in regards to microdata will most likely be the straw that breaks the camel’s back. It will discredit the HTML5 effort, and will serve to taint its usefulness into the future.

    The best thing about all of this is that the more you try to assert control over HTML5, the less control you will end up having.

  21. Bijan, “As a final point (while the flu kicks back in :)), I’m not sure what’s *wrong* with harvesting comments from blog posts. Who cares if they aren’t intended as a formal submission? Shouldn’t we actually look and see if they are chaff?”

    Now there speaks someone who has, most likely, also never participated in the requirements process for a business application. Going out and gathering requirements from casual comments is little different than grabbing a pad and pen and hanging around the water cooler, and using this as the primary basis of designing a requirements document.

    Frankly? Ian should never have been the person managing all of this. I don’t understand why Sam Ruby and Chris Wilson continue with this farce. Why does the editor of this document have so much power over every last aspect of HTML5? I still am astonished that the W3C would condone this. My respect for the organization has really tanked with this effort.

  22. @John Foliot: you appear to be labouring under the misapprehension that you (and/or people who agree with you on particular issues) are the “the Accessibility Community.”

    Some folks in the accessibility community are happy with the way HTML5 is addressing accessibility, and some aren’t. There’s a range of opinion on alt, longdesc, summary etc among folks who evangelise accessibility, and the community benefits from this diversity, imho.

    Neither you, nor any other single individual, represents “the accessibility community”. Try remembering this please.

  23. @Shelley:

    See RDF is a great technology for certain fields — Hixie.

    Also, how do you ensure comprehensiveness of given use-cases? How do we make sure everyone who gives a use-case on your blog formally submits it to the WG? If we do not, it will be less complete than it does otherwise. The WG will have to vet the use-cases anyway, so what differences does it make where they come from?

  24. @Shelley. “Kyle’s opinions are not discredited, just because Ian comes in and say”. Well, Kyle retracted, so presumably he thinks otherwise. You’ve now claimed that a direct report from a person of what they did carries less weight than speculation about what he did.

    If you really doubted Ian, why not just ask for the text that was the basis of his derivation? You could then examine them and Ian’s result and make some judgement about his ability to select and present use cases. That would be much better than the insinuation that he’s a liar.

    @Shelley. “Now there speaks someone who has, most likely, also never participated in the requirements process for a business application.” I’m not sure if my experiences “count” with you (would doing so for NASA count?), but I don’t see how it’s relevant anyway. But let’s dig in.

    First, I don’t see why even water cooler talk isn’t a useful source of *data*. If users of an app are complaining about some feature, then it doesn’t matter if they’ve “formally submitted” or not. Of course, you don’t take gripes at face value, but I wouldn’t take “formally submitted” comments at face value either. (Esp. as “formally submitted” is, at the moment, undefined.)

    To put it simply, there’s no evidence that Ian took the blog posts comments as anything other than data for distilling use cases and requirements. That seems totally normal and appropriate. He then sent his distilled set of use cases and requirements to the WG mailing list. Which, again, is pretty normal and, I think, reasonable. You may question whether he did a good job in that or not. But it’s rather strange to claim that he’s somehow done something bizarre or out of bounds or unprofessional in that. (Note that Kyle’s point is different, to wit, that the blog post comments are likely to be a poor source of data. That can be checked, of course. But then it becomes a time management issues. If someone *can* scan through the chaff to find the wheat and is willing to…what’s the problem?)

    Second, I have participated in a lot of W3C groups and spent several years specifically trying to build up a community based set of requirements for a W3C spec (OWL 2). Frankly, the kind of thing Ian’s doing doesn’t seem different in kind. For example, every WG has a public comments list and, indeed, has to take (esp. during LC) all public comments seriously. There’s *no* vetting of the comments other than they need to be mailed to a list in a certain period. Anyone can send them. You have to deal with all of them. Given that, it just makes sense to go looking for comment from interested parties or, at least, data about what they want or believe they want.

    So, I don’t see that your criticisms are on point. I can see critiquing the substantive outcome (i.e., you could claim Ian did a bad job of distilling requirements and uses cases from the data, that he neglected key data, or both; but then you’d have to show that substantively). I can see critiquing the methodology (e.g., on efficiency grounds or, as I think Kyle has tried to do, on whether the methodology will support stakeholder buy-in). But I’m not sure how effective it is to claim that the outcome is bad *because* the methodology is un-specifically “inappropriate”. (Similarly, pure methodological defenses can’t make outcomes correct. For example, a methodologically impeccable survey might still be wrong!)

    BTW, could you point me to some places, if you have them handy, where Ian rejects “formally submitted” use cases (perhaps this just means use cases identified as such by the submitter specifically for consideration in the design of HTML 5?) on ill-formedness grounds without giving useful criteria for well-formedness? (I’m not saying that it didn’t happen, but I like to check these things out myself.)

  25. Geoffrey, no one “gave a use case” in my comments. They’re comments, for goodness sakes. And I haven’t a clue which ones Ian mined or didn’t. For all I know, he could have taken as gospel truth something someone wrote in satire…because they’re comments!

    No, not public comments on the spec to the W3C — ones Ian found in my weblog.

    As for pointing out the one IRC that says something relatively positive about RDF, I can point out several in the last few days where Ian has been less than complimentary about RDF.

    Bijan, no, I don’t agree that what Ian did was best. Evidently, you do. As for judging whether he’s done a good job of condensing the use cases down or not, how can I tell? Do I have access to the emails, comments, etc he’s used?

    And now, I’ve seen him take one of the use cases, and in the WhatWG list, spell out his reasoning why it’s not a valid use case. He’s already decided, without any form of discussion. The only discussion seems to be, email him with your thoughts. Don’t take my word, read it yourself.

    This isn’t some cute little hacky markup you’re all putting together. This isn’t scientific research. This is supposed to be the next version of HTML — this is massively important. Too important to have one person make all the decisions. I don’t know how you all can think otherwise.

    Obviously, you do, though. Great, good for you. But it’s not good for me. I no longer trust what comes out of the HTML5 effort. I have no faith that it is the best effort.

    This has all boiled down to our trust in one individual. This isn’t a team effort, or consensus, or ensuring comprehensive coverage of all interests. This is all hinging on our faith in one individual. That should be enough to alarm everyone, regardless of how you feel about the person.

  26. @Bijan – As you mentioned farther up, perception and PR do matter to the success of the spec. At present I believe this isn’t being managed, so I’m not expecting much success in the process going smoothly at Last Call. Regarding Mark’s comments specifically, as a major contributor to the WhatWG Blog and frequent conversationalist in the public WhatWG IRC channel, his voice gets attached to the HTML5 efforts regardless of the accuracy of that connection. I don’t think he needs a time-out, or some sort of exile to a far land, but at the very least the actual WhatWG members should simply speak up with a retort to his comments to the effect of “Hey, personal insults aren’t cool, aren’t professional, and are hampering the process. Cut it out.” The fact that it seems unreasonable to you that I suggest professional behavior from a body of individuals writing an international web standard is startling to me. Even more startling is how begrudgingly people have been about label his comments as inappropriate when their attention is called to them.

    Shelley’s comments reflect my opinion on the evidence of bias regarding Ian’s consideration of RDFa. If you feel that these examples somehow fail to illustrate the issue, then we can simply agree to disagree. Regardless of whether there is bias, there’s been a failure to help prevent that perception, which once again will hamper the overall HTML5 effort.

    Regarding Ian’s bias or lack thereof with the spec: Ultimately, what this comes down to is that I lack the ability to trust any one individual with the future of HTML5. There’s simply too much at stake and too much room for error. Ian may be a great guy. That’s not the issue at hand as to whether he’s capable of writing a good spec. The question is whether the results will suitably reflect all interests involved (or have at least fairly addressed them.) At least with an attempt at genuine consensus, you can show that the effort was made. At present, I believe there’s a genuine lack of evidence that bias isn’t playing a part here, consciously or otherwise.

  27. Shelley asked for the sources from which I compiled the use case list, so I sent them to the www-archive mailing list, in case anyone is interested:

  28. @Shelley.”Bijan, no, I don’t agree that what Ian did was best. Evidently, you do.” What is the evidence that I do? All I’ve argued is that your criticisms, thus far, as I am able to understand them as you’ve expressed them here, don’t seem to be on point. I still don’t see, prima facie, why taking blog post comments as input is so far out of the bounds of reasonable behavior as to suggest to you that my not seeing that they are out of bounds indicates that I have no experience with requirement analysis.

    Please don’t attribute views to me that I’ve not actually expressed (or, for that matter, even implied!) It’s one of my pet peeves. Thanks!

    I see you are asking for the data from which Ian derived his use cases (on the HTML WG mailing list). I’ll be interested to see what comes of that. Of course, in the end, the use cases will have to be evaluated on their own merit as well.

    @Kyle. First, Mark Pilgrim is definitely *not* a “frequent conversationalist on the public WhatWG IRC channel”. At least, not in my observation. (Just correcting a fact there. Your point is probably untouched regardless of whether he is or not.)

    “”"The fact that it seems unreasonable to you that I suggest professional behavior from a body of individuals writing an international web standard is startling to me.”"”

    I don’t know that it is required of professional behavior to take note and/or publicly censure someone for off hand comments made in a chat room. I’m not even sure that it is required of (overall) professional behavior not to make off hand comments in a chat room. People say all sorts of things all the time. People vent all the time. Chat and email make things that were once ephemera and private public and permanent. Some people think that means we should behave as if we are to be on our best behavior all the time. Other people think we must try to direct our attention differentially.

    (This is different as to whether it is *productive* to make off hand comments in a chat room.)

    For example, it’s not uncommon to regard public policing as being unprofessional as well. (E.g., on many mailing lists, it’s the norm to complain about bad behavior *privately*.)

    Personally, I tend to regard, as I said earlier, sober sounding accusations of unprofessionalism much more serious than over the top personal insult. After all, the latter tends to discredit itself and the speaker (at least, it tends to put readers on their guard). The former tends to be accepted and thus can damage the reputation of the target of the comment.

    In any case, I’m fine saying that Pilgrim’s comments were unhelpful. The fact of this conversation shows that.

    Finally, I’m still confused by the mix of bias accusations and “generating a perception of bias” accusations. I tend to regard the former as primary, so if you would explain again to me the evidence of bias about RDFa, i.e., its existence, it’s influence, and the problem with both (without regard to perception issues, which I think are separate), I would be grateful.

    As for trusting “any one individual” with the future of HTML5, I don’t see that anyone (I hope!) is advocating that. Ian is a focal player, clearly, but it’s not like he’s operating in a vacuum. It’s very unclear to me (as a close watcher of the HTML5 WG) *what’s* going to happen. Ruby (who has more formal power in the group than Ian and a lot of support in and out of the W3C) seems to be angling (in my opinion) for a more minimal version of HTML5 wherein every part of it has consensus (by his standards). He seems rather sympathetic to RDFa and distributed extensibility. Rob Sayre’s draft of the spec is very different than Ian’s (but I don’t know if he’d support RDFa). It’s not even clear to me that Ian won’t include RDFa, in the end, in his draft. I don’t know that that’s the best outcome. (BTW, Ian is hardly the only person who has expressed antipathy to RDFa (or some of the design choices). See . And it’s easy to find Roy bashing Ian :))

  29. I guess I should make clear that I *don’t* have a substantive opinion on most of HTML5. I like the goal of specced error handling a lot. I read through the parsing spec a while back in preparation for a class. It wasn’t particularly easy going for me, but without trying to implement it or write test cases (or spending a lot more time with it), it’s hard to judge the actual clarity of the text or the structure. I am a semantic web person, but I’m not personally enthused by RDFa and I don’t know whether it’s wider adoption or inclusion in HTML5 would be good for semantic web effort (at least the parts of more direct interest to me).

    I do care quite a bit about process at the W3C (since I regularly try to get specs through there) and I care about general professional standards and about ethical behavior. Hence the commentary.

  30. This exchange between Shelley and Ian e.g.,:

    looks quite promising. I’m particularly impressed at Shelley’s engagement and tone (given her expressed frustration with some of the players and aspects of the situation). Good show and I hope good things come out of it.

  31. Hi Kyle, I found this post via the “tantek-celik” tag on your HTML5 Super Friends post.

    Apparently I was involved in a bunch of drama and didn’t know it? In my experience most of these sorts of drama/politics claims tend to both be inaccurate and not very productive. Some of them border on trolling, and thus I tend to ignore them in order to focus on more positive, productive ends.

    To address a couple of factual points in your blog post:

    1. Ian asked me for “feedback” (per the IRC links in your post) on use-cases – not vetting. I know the terminology distinction is minor, feedback on many aspects of HTML5 is strongly openly encouraged by a broad range of people, whereas “vetting” sounds like some sort of qualification process. I gave feedback, and did not perform any sort of “vetting/qualification” role.

    2. neither Ian Hickson nor I “attack Creative Commons”.

    I’m fairly well known for being a longtime strong supporter of Creative Commons the organization, and have been publicly Creative Commons licensing my presentations for years.

    What I have criticized is:

    a. the original Creative Commons RDF-in-HTML comments technology which nearly no-one used and was quickly superceded/dwarfed by use of the much simpler rel-license microformat in 2004.

    b. more recently, as noted in the IRC archive, ccREL – for having LOTS of issues, too many to document here, but in particular, being contrary in design to one of the goals of Creative Commons, which is to reduce license proliferation. More details here:

    Creative Commons is a great organization. Unfortunately they’ve had mis-steps when it has come to inventing technology to serve their ends (mostly because of the unnecessary complexity and invisible data flaws which come with using various flavors of RDF syntaxes).

    One can very much (and should!) criticize any such technology when it fails to support the higher-level goals and objectives of the organization.

  32. @Tantek – I had to re-read this post to remind myself of why I wrote it. That may say something about a developing internal sense of the HTML5 process and microdata “debate”.

    Let me apologize for the tone in which this was written. At the time I was mad at what Mark Pilgrim was saying (I think it’s fair to say he’s known for some crude commentary at times) and ended up viewing other conversations in a less than charitable light as a result.

    In reviewing this now, I agree that you had no ill intentions or unduly unkind comments for me to criticize you for. In particular, I misconstrued your Creative Comments statements in a broad-blanket fashion that was unfair.

    Thank you for noting the errors in what I construed, and providing your actual criticisms. I agree that criticism can and should happen.

  33. Thanks Kyle, I appreciate the update. FWIW I think there are issues with microdata as well. Though simpler/easier than RDFa, microdata is still far too complex for even experienced web designers, and the predefined vocabularies are (ever more so) obsolete snapshots of hCard and hCalendar which are being actively maintained (faqs, issues, resolutions, errata) on I’m working on 1.0.1 updates to hCard and hCalendar which should remove any need for duplicating vcard/vevent vocabulary in HTML5.