Help keep accessibility and semantics in HTML

This is a call to action directed at all standardistas and accessibilitistas. If you think accessibility and semantics are important and should be improved in the next version of HTML, you need to act.

What is currently going on in the W3C HTML Working Group is very disappointing and something I never expected to see when I joined it. I was naive enough to think that everybody joining the HTML WG would be doing so out of a desire to improve the Web. Unfortunately, that does not seem to be the case:

  • There are people who argue against the value of semantics and in favour of keeping or even adding more presentational markup.
  • In the HTML 5/Web Applications 1.0 Working Draft, accessibility features such as the summary, headers, and axis attributes (which are used to make data tables more accessible) have been removed with no explanation.
  • The general knowledge of accessibility in the Working Group is surprisingly limited.
  • Some of the loudest voices are browser vendors who seem to have an extreme phobia of doing anything to discourage the use of invalid and inaccessible tag soup.

All in all, my impression so far is that unless things change, the next version of HTML will do nothing to improve the Web. All it will do is make things easier for browser vendors and ignorant web developers. The rest of us may be better off sticking to HTML 4.01 Strict.

We need to change that if there is to be any chance of cleaning up the mess that the Web currently is.

I do what little I can to argue, but it takes more time than I have at my disposal. I also lack the rhetorical (and English) skills needed to respond to the dismissive and belittling arguments that meet anyone who opposes the majority view.

If you have an interest in improving the accessibility of HTML, want more semantic and less presentational markup, and are good at arguing your case, please consider applying for HTML Working Group membership by following the Instructions for joining the HTML Working Group. Do it now. The longer you wait, the harder it will be to change the unfortunate direction things are going in.

Thanks.

Posted on May 7, 2007 in (X)HTML, Accessibility, HTML 5, Web Standards

Comments

  1. May 7, 2007 by Tom

    That is very, very disappointing to hear :(

    Sadly, as a student I think I am unable to join and argue my case.

    I really hope that some sense can be brought to the discussion and that we will not take a step backwards with HTML 5.

  2. i’ve been waiting for a week now my application to participate in the WG to be approved…but from what i’ve spied so far from the web archive of the list, i’m seriously worried…

  3. I’m with you my brother and people are discussing this topic on my forum as I type this.

  4. May 7, 2007 by Chris Boudy

    Roger, you can’t be serious?

    It’s like I woke up in the twilight zone. I hope these are not “web professionals” who are saying that stuff.

    I’ll definitely be looking into joining now!

  5. Hey there Roger,

    I have seen some of your posts on the list and I am sooo… glad that there is someone else that feel the way myself and my colleague Frank feels.

    A lot of the discussions are so convoluted and pointless but still, they seem to carry on forever. I don’t see a lot of discussion about improving the current HTML language and any attempt to open a discussion is quickly snubbed out.

    I hope we can get some more people on this list and that we can unite together to do what I, and you yourself hoped to achieve, improve HTML for everyone.

    Bugger the browser vendors who continually throw their weight around and stand in the way of a beneficial move forward.

    See you on the list!

  6. I’ve just finished reading new WCAG 2.0 draft and can’t understand why these groups (WAI and HTML working Group) are moving in completely different directions?

  7. This is very worrying indeed. I don’t know how much I will be able to contribute (or if I will even be accepted), but I have begun the process of becoming an “Invited Expert” (which seems to be unnecessarily complicated IMO).

    I hope we can do something about this.

  8. I just looked into joining. I was disappointed to see to process is like many thing W3C: long and drawn out. I’m almost afraid to get into this thinking it’ll dominate my life. I’m torn.

  9. This is disgraceful. The W3C are in real danger of going even further down a route to irrelevance. This just compounds their inability to output relevant and finalized standards, their endless delays (CSS3 anyone?), their split visions (whither XHTML2, and why is it not backward compatible?). If the governing body of the Web can not stand up for its own ideals, has lost focus of why it exists and who it’s supposed to be serving (end users, not browser vendors) - then what is the point in the W3c? Were the last 6yrs+ for nothing?

    Distressing news, and more so due to the archaic nature of the W3C and the obscurity these important decisions are shrouded in. Mailing lists? In 2007? Who has the time and disposition to use such archaic means of communication in an era of search-able, instant answer, easily reviewed forums.

    A lot of the recent rumblings about and from the W3C are causing me to seriously question the future of the web and whether I want to be a part of it. It’s not going where I thought it was going. As a dedicated designer and developer who’s top priority has been universal access for all, of interoperability, of doing things right it’s depressing to watch the people who make the rules flop about like some sort of impotent and listless dinosaur, contradicting themselves and generally getting trapped in a mud flat only to await extinction, while other bodies evolve to fill the gaps.

  10. I’ve applied to become a member. A process as convoluted and confusing as the rest of the W3C (The highlight for me being that I had to look up what the heck an ‘international phone number format’ actually was, then convert my number into it). I hope I pass the grade.

  11. Please help me: Don’t understand why HTML can’t just be semantics.. Didn’t we speak about separating structure, content and design for a long time? Wasn’t there an approach to use HTML only to mark a text up - to structure it by it’s original semantics? Can’t remember, but i thought you’ve to use “style”sheets for presentation purposes?

    Hope those guys in redmond, oslo and all over the world won’t support this….

  12. Roger, I think you are misrepresenting the issue here. The public-html mailing list has gotten out of control to the point where meaningful discussion is impossible. I don’t think anyone is arguing that semantic elements should be eliminated or discouraged— although there are some (myself included) that are advocating the retention of presentational elements in addition to the semantic ones for a number of reasons (I will write a post about this on my blog). This will not prohibit anyone from using semantic elements, and it is not implied that presentational elements will be a part of conforming documents.

    As to the missing table attributes, I thought it was made clear that they were not removed, but instead that they had not been added yet. The WHATWG spec is incomplete. I’d bet good money on the fact that they’ll be in there when it’s all said and done, especially since there is demonstrable, real-world benefit to their inclusion.

  13. Roger,

    I totally agree with you. For the past couple weeks I have been sitting back and watching people try and make a case for such things as BOLD and ITALIC tags. Being new to the group I haven’t felt the justification for me to respond with my thoughts just yet, but I think that is going to change. Standards evangelists like yourself and countless others have spent way to many hours to “not hurt the web”. I believe that dedication needs to be acknowledged and those practices be placed into HTML 5, not things such as INDENT tags and stripping out of accessibility features.

    What can we do as a group (members and non-members of the working group) to assist in making the web a better place, not a place of tag soup, nested tables and non-semantic code?

  14. The current development of the HTML5 and XHTML2 specs has my mind going in circles. Which to choose? Which to support?

    Either way, here is what my good friend spikku is recommending:

    “Why isn’t there a <nav> tag? We need a <nav> tag!”

    I fully agree. Lists are nice, but it would be even nicer to have a tag which we could use for the global navigation of a website. <nav> FTW!

  15. Twillight zone indeed.

    documents that use the noscript feature can be represented using “HTML5”, but cannot be represented with “XHTML5” and “DOM5 HTML”

    I.e. the noscript element is banned from the XHTML5 serialization of the (X)HTML5 markup language in the current draft of the spec…

  16. The HTML5 spec remains a work-in-progress. I do not believe that any accessibility features/elements have been deleted; I don’t remember any conversation on The WHAT WG mailing lists or IRC to that effect.

    Invalid, nonforming content is a different issue. Browser manufacturers have been (and, continue) displaying invalid code in “text/html” web pages. That won’t ever cease. Just as web standards education should not cease.

    There seems to be two advocacy camps: graceful error handling and hard [XML-like] error handling. I have not seen any proposals for an inbetween method. Maybe, an invalid CSS-like method; or, IE6-sans-hacks method.

    Semantics discussions on the mailing list get murky.

    I do not believe that the HTML5 spec will harm Web Standards. Especially, since it is a working-draft-in-progress.

  17. Hmm … am going to have to chew on this one for a bit, to see if I have the required energy to get into the action.

    With regards to “graceful” vs “hard” error handling, I’d like to see something in between the current implementations of very forgiving browsers, and very unforgiving xml parsing. But then again, I’d also like to see DOM and CSS implemented properly across the board.

    I have very deep respect for Chris Wilson, but still find it disturbing that he’s gotten such an important chair, considering his employer and their trackrecord when it comes to respecting any standards. Chris might have the best of intentions, but he’s not just representing himself, and will have to keep his employers intentions in mind, otherwise he’s not really doing his job, is he?

    </random-thoughts>

  18. As a fellow member of the working group, I’m as worried as you are, Roger. I wasn’t expecting this kind of stuff from any of the members of the WG. I joined the WG for one reason: To participate in the making of a more accessible and standardized HTML version. I don’t want more presentational elements in the spec (though I can’t see the problem of keeping BLOCKQUOTE, to keep the “web from breaking”). I must say I’m very disapointed in many of the browser vendors and their view on how the new version of HTML should be. It’s a MARKUP-language for crying out loud, and should have as few presentational elements as possible, and I really can’t see why the browser vendors and other people worried about accessibility, semantics and standards can’t feel the same!

    I used to be “in to” reading as many of the discussions on the mailing list as possible, but now I just mark the ones I find interesting, read those, and mark the rest as read, simply because there are so many people on the list that either don’t have a clue, or they’re just ignorant. I don’t have the time or necessary skill to argue either, but that could just be bad self-esteem.

  19. May 7, 2007 by Roger Johansson (Author comment)

    Jeff:

    Roger, I think you are misrepresenting the issue here. The public-html mailing list has gotten out of control to the point where meaningful discussion is impossible.

    Misrepresenting, misunderstanding, whatever. I hope I am completely wrong and none of my fears come true. But this is the way I feel about the discussions on the public-html list. And yes, meaningful discussion is completely impossible.

    I don’t think anyone is arguing that semantic elements should be eliminated or discouraged

    It depends on how you interpret their arguments. The politically correct opinion seems to be that anything that is widely used should be adopted (“pave the cowpaths”). And that leaves little or no room for semantics and accessibility.

    As to the missing table attributes, I thought it was made clear that they were not removed, but instead that they had not been added yet.

    I certainly hope they will be added then.

    Doug:

    What can we do as a group (members and non-members of the working group) to assist in making the web a better place, not a place of tag soup, nested tables and non-semantic code?

    • Insist that tag soup and bad practices are never ever mentioned anywhere in the spec unless they are accompanied by a big warning sign saying “Do not use!”.
    • Insist that error handling for browsers is mentioned far away from the parts of the spec that web developers will read.
    • Insist that browser vendors implement some kind of error logging for HTML, like iCab does.
    • Insist that HTML 5 (if that will be its name) is a real improvement for standards and accessibility aware designers and developers, and not only for browser vendors.

    Mike:

    Why isn’t there a <nav> tag? We need a <nav> tag!

    There is one in the current Working Draft :-).

    Sean:

    There seems to be two advocacy camps: graceful error handling and hard [XML-like] error handling. I have not seen any proposals for an inbetween method.

    My opinion is somewhere inbetween. Though I would like draconian error handling, I realise that will never happen. So instead I think browsers should handle errors, render the page, and somehow make users aware of pages that contain errors. As I mentioned previously I like the iCab way of handling that.

    I do not believe that the HTML5 spec will harm Web Standards.

    I hope it won’t.

  20. “I have very deep respect for Chris Wilson, but still find it disturbing that he’s gotten such an important chair, considering his employer and their trackrecord when it comes to respecting any standards.”

    I think it’s thinking like that that led to Dan Connolly being made a co-chair alongside Chris. Fair enough, I think.

    I currently have about 300 unread working group emails at the moment, and new ones arrive faster than I can read the old ones. Everybody appears to be arguing about something, but I’ve lost track as to exactly what, and suspect I’ll be leaving soon, it’s just too much.

  21. May 8, 2007 by Scott

    Once in a while I’ve mulled over whether to join. Initially I decided against it because I thought there would be plenty of smart people who know better than me anyway and will do a good job. Now I’m still deciding against it because it doesn’t sound like it will be possible to make myself heard, and I have enough stress in my life without worrying about HTML 5.

    I agree that mailing lists aren’t going to cut it this time. Forums would definitely work a little better, divided up one forum for each module. Even then, it would probably still be chaotic.

  22. Roger,

    When you mention that:

    Some of the loudest voices are browser vendors who seem to have an extreme phobia of doing anything to discourage the use of invalid and inaccessible tag soup

    Can you be more specific? I’m baffled why anyone but Microsoft would have a vested interest in being so backwards in approach. Gecko and KHTML/WebKit have both benefitted greatly by adopting semantic standards set up with HTML 4.01 and XHTML 1.0-1.1. How would they see stepping backwards as an advantage to anyone but Microsoft who still falls far short of supporting seven year old standards?

    Now that both iCab and Opera can claim Acid 2 passing grades, why would they want to go backwards either?

  23. PS: I too like the way iCab handles alerting the end user as to the quality of the viewed page’s code. Same goes for when I use tidy plugins with Safari or Firefox.

    I use it as a minor indicator as to the trustworthiness, competence or reliability of a company whose site I’m visiting.

    I think that if this could be sold to browser vendors as a win for end users they might start thinking differently.

    Companies and web developers collectively and as a whole would start to see the value of better-formed code. They’d see the public relations/percetion benefit of delivering “green light” code over amber or red.

  24. Paul haine: I would suggest that you wait until asked to leave. Participation in the Surveys remain very important and your WG Member vote remains very important.

    John Lascurettes, Roger: I am inbetween, too. However, my thoughts are much less generous than yours. My pipe-dream is to have error handling be performed in a similar manner to IE’s handing of CSS box model and undimensioned floats. Technically, it would not “Break the Web”; invalid pages would show design content (but just not as the author intended).

    iCab is a good compromise but the general public won’t know, won’t care.

    Web Standards Life would have been so much simpler if draconian error handling were introduced in HTML 2.

  25. May 8, 2007 by Tommy Olsson

    If I thought for a minute that it would make any difference, I might join. But from what I’ve heard, it seems at least as bad as the scenario you described.

    IIRC, Simon Pieters even said once that they were redefining the meaning of ‘paragraph’ to include all sorts of block level containers. Apparently some people thought it was ‘unsemantic’ to wrap form controls in DIVs, and thought that P elements would be better.

  26. May 8, 2007 by Roger Johansson (Author comment)

    John:

    Can you be more specific? I’m baffled why anyone but Microsoft would have a vested interest in being so backwards in approach.

    It’s not so much that they have a backwards approach - they do want to keep improving their support for web standards, which I believe even Microsoft wants. The problem is them insisting on not doing anything to let the user know that there is a problem with a page, in fear of losing market share. At least that’s the way I interpret what their representatives have stated.

    paul: I have also considered leaving, but as Sean says I think we should stick around for a while.

    Sean:

    Web Standards Life would have been so much simpler if draconian error handling were introduced in HTML 2.

    Yep.

    Tommy: I really do think someone like you could help make a difference.

  27. Roger, sorry if this is a repeat of a previous comment, I am answering in a hurry and for good reason; I am too damn busy with work and family to consider joining a working group. I am busy enough battling FOR reasonable HTML in my day to day work let alone arguing to save it outside of office hours.

    What about a petition for us to sign? Listing elements that need saving and points of view that need weight? One person with the support of 100s maybe 1000s could be a more powerful than an uncoordinated group?

  28. Why isn’t there a tag? We need a tag!

    There is one in the current Working Draft :-).

    Sticking to semantics, I still view a navigation as some sort of unordered list. There is the role attribute to add semantics like role="navigation" or role="breadcrumbs". I’m strongly against any segregation of perfectly suited elements into multiple others (or too many, like abbr and acronym).

  29. Here’s another agreeing vote.

    The new spec sounds like “If we make theft legal, crime rates will drop”

    I much preferred what was being suggested for XHTML2, but with something more like HTML5’s error handling. I understand this means it wouldn’t be xml-based, but it’s only a suggestion.

  30. May 8, 2007 by Darren

    Heres easier instructions for joining the working group (it took me less than a week):

    http://ln.hixie.ch/?start=1173385976&count=1

    Can we get some more heavyweights talking about this issue - Molly, Zeldman, Eric Meyer, amongst others…

  31. Wow, there are still people who think that draconian error handling is a good idea? If you do, I suggest you take it up with your elected representative. Because without regulation in the form of law, in an unregulated Web browser market, draconian error handling will never happen regardless of what any specification says. Any Web browser that dared to implement it would be abandoned by its user base, in favor of browsers that displayed more Web pages.

    Even if by some miracle browser vendors had agreed to draconian error handling for HTML back at version 2 (as Sean and Roger wish), HTML’s unreliability would have made it highly unpopular with authors, and it would have been replaced by something else.

    (Both of these phenomena are demonstrated in the unpopularity of XHTML-served-as-XML, amongst both browser vendors and users.)

    CSS’s error handling is not draconian in the way that XML1’s is. When a browser encounters an error in a CSS rule, it doesn’t throw out the entire style sheet, it throws out just that rule. It would have been nice if HTML had had error handling as simple as that specified many years ago. But because it didn’t, HTML5’s error handling has to be more complex now.

  32. Roger, I’m very surprised by your comments, but I’d like to address some apparent misunderstandings.

    There are people who argue against the value of semantics and in favour of keeping or even adding more presentational markup.

    That is not true. We are against semantics for the sake of semantics. In other words, adding a new semantic feature actually has to have:

    • Real use cases to explain why authors would use it. If there’s no reason to use it, authors won’t.
    • Solve real problems. We need to identify the limitations with existing HTML, figure out if existing HTML can already solve the problem; and if not, why not? If existing HTML can’t solve a problem, then we can consider finding a solution.

    • Provide some incentive for authors to use it. A feature in the spec that won’t be used is no better than a feature not included at all. So any proposed feature needs to show that there will be some real incentive for authors to use it.

    No-one sensible is advocating the addition of even more presentational markup. Retaining b and i is a pragmatic one, and frankly the whole debate turned into a bike shed. It is simply impractical to provide semantic elements for every possible semantic use of bold and italic, and there would be no benefit in doing so anyway.

    There is no harm in providing catch-all elements for the remaining use cases. In fact, there is even some benefit in using those presentational elements for the cases where HTML lacks specific semantics, because it at least indicates something that can be determined from the context.

    Those arguing against it are simply arguing on the grounds that a strict separation of presentation and content is a necessary goal in itself. It’s not. It’s a means to an end, and if that end can be reached without a strict separation, then so be it.

    In the HTML 5/Web Applications 1.0 Working Draft, accessibility features such as the summary, headers, and axis attributes (which are used to make data tables more accessible) have been removed with no explanation.

    They haven’t been removed. They just have not been added yet due to lack of evidence to support them.

    The headers attribute: I’m aware that this one currently has better support in ATs than the scope attribute does, but for most cases it’s redundant.

    If the problem is just associating cells with their headers, we should investigate alternatives that would make it easier, such as defining an algorithm for more accurate implicit association.

    That would be better because it increases accessibility, while reducing the requirements on authors. However, that needs research and evidence to determine if it can cover sufficient use cases reliably, which will allow us to figure out if headers is still required.

    The axis attribute: Very few people understand what this is for, and even fewer actually use it. What problem does it solve? What are the use cases? What’s the incentive to use it?

    The summary attribute: This one just needs more research to determine if it’s worthwhile including. There are stats to show that the attribute is used relatively frequently, but those stats don’t indicate the signal-to-noise ratio for it.

    If research shows that it’s far more widely abused, than used usefully, there’s little benefit in retaining it. Conversely, if research shows that when it is used, it’s often used correctly, then it would be good to add it.

    The general knowledge of accessibility in the Working Group is surprisingly limited.

    People from the accessibility community have been welcome to join in the WHATWG for years, but haven’t. It’s great that more of them are joining the HTMLWG now.

    They just need to understand that they need to make their case very clearly and respond to questions properly, not just complain when someone provides counter arguments. (This is not aimed at anyone in particular, but there are a few people who behave like that).

    Proposing solutions like the role attribute and RDF without clearly identifying what problems it really solves and why those problems can’t be solved with existing markup, doesn’t help. Personally, I don’t think role is a good solution to anything. It’s over engineered and far too poorly defined.

    Some of the loudest voices are browser vendors who seem to have an extreme phobia of doing anything to discourage the use of invalid and inaccessible tag soup.

    Encouraging authors to do the right thing is a far more effective and less harmful than trying to discourage (or worse, punish) them from doing the wrong thing. Browser vendors simply cannot do anything that will break the web or harm the potential for HTML5 to be adopted.

    There seem to be a lot of people who have the wrong impression about the group. With a group this size, not everyone is going to get their way. We need well thought out, reasoned discussions. From my perspective, there is a lot of push of ideologies that are not supported by reason or logic, yet do have reasoned and logical arguments against them.

    People just need to calm down, and stop over reacting to the situation. The spec is still a draft and a long way from completion. I’ve noticed some people seem to have the opinion that because the spec includes something they don’t like, or omits something they do, that the spec is broken by design. Some people use that as the basis for wanting to throw the whole thing out. That is not the right approach. A better approach is to identify problems, ask questions, make your case with reasoned and logical arguments, and, most importantly, listen to counter arguments.

  33. May 8, 2007 by Stevie D

    With regards to “graceful” vs “hard” error handling, I’d like to see something in between the current implementations of very forgiving browsers, and very unforgiving xml parsing. But then again, I’d also like to see DOM and CSS implemented properly across the board.

    This is where versioning is important.

    Any page that claims to be HTML4.01 or less should be rendered as it is now - no “policing” of invalid pages, because this would completely break the web.

    Any page that claims to be HTML5 can/could/should be rendered with tighter regard for the specification. I’m not saying a complete XHTML-style stop, but a degradation in presentation, more akin to turning off CSS.

    Yes, some errors would pass unnoticed; there would be no need for the browser to make a fuss. But others would cause the browser to react in some way. Not (usually) to the extent that the page is unusable, but to remove any possibility of the browser making an incorrect interpretation of ambiguous code.

  34. May 8, 2007 by Tom

    This thread is the perfect example why after 12 years I am so disillusioned with this part of my craft. Where we are now and where I thought we would be are such vastly different places.

    First of all, it’s laughable - laughable - that they are using a mailing list for this group. How is that effective? But I even question if this type of group is the right way to go about getting HTML moving forward, because I think we already see the problems inherent in the model.

    Someone mentioned the irrelevance of the W3C and I couldn’t agree more. Where is the promise of innovation and invention in this space that we hoped for over a decade ago? Is anyone disappointed as I am that we’re using decade-old technology in a field that has seen historical explosive growth?

    And we wonder why we have tag soup. Stop wondering.

    In the beginning we had the Browser Wars(TM) and we thought that was a Bad Thing(R). But you know, the browser wars brought forth lots of innovation, didn’t it? Sure, it gave us more tag soup. But it seems to me that innovation in HTML got stifled the moment the W3C and the well-intentioned Web Standards Group got involved.

    Standards have brought us lots of Good Things(TM) - and I am a proponent of them. I’m not saying we should not have done some of the things that have been done. But we have clearly derailed and lost our way.

    Where is the innovation today? It doesn’t come from any “working group” (that never seems to get any WORK done - apologies to those of you who I know are trying hard to do just that). It comes from people in the “Web 2.0” world who have taken things into their own hands with the technology that is available to do new and exciting things - in SPITE of bad hand we’ve been dealt.

    I can hear it in so many comments here - the weariness of people in dealing with this issue - especially from those of you that have been around awhile. Like Roger says - who has time for this? I don’t. We have families to support and lives to live.

    What’s the answer? I don’t know. It’s not this working group though. If and when they finally release something (which the W3C failed to do, essentially) that actually gets implemented, it will be a LOOONG time from now - and by that time, it may not matter anymore. Because innovation on the web cannot wait.

    I realize this is a cynical rant, but I also realize I’m not the only one who feels this way. I don’t know where this thing goes from here, but I’m tired of it, and I’m not hopeful it goes anywhere useful. I expect the real innovators in the trenches will continue to use web technology in interesting ways, again, in spite of what comes out of these working groups. And standards and accessibility advocates, like Roger, will have to continue to simply work with the tricycle we all ride that is HTML.

  35. May 8, 2007 by Matthew Smith

    This is why the W3C and future versions of HTML are so unimportant. Rather than dealing with important issues like how to make HTML more feature-rich so it can be a better rich application platform, we are stuck with arguments on semantics and accessibility. Maybe you should focus discussions on how to release a new version of HTML and CSS more then once a decade.

  36. I just want to say that I thoroughly support all of you who are attempting to work with the HTML Working Group. It sounds like a difficult task so I hope you’ll keep at it.

    Unfortunately for myself, I’m not very good at arguing with people who don’t get it. Well, I’m pretty good at forming an argument but obstinancy on the other end drives me insane. Not good for my mental health!

    I did mention this to my husband and he is interested - he has looked at joining the group before but he was thinking they only want people who work for big companies. It seems like it is possible for independent people to join the group - are there any qualifications to become an “invited expert”? I poked around on the W3C site a bit and it seems that it’s mostly got to do with who your employer is and why they’re not a W3C member.

  37. Joining the W3C as a company/organization, is something quite expensive.

    Even for non-profits the annual fee is like 2000 EUR for some countries (!) and for a small company it is 6500 EUR (!!).

    For other countries, non-profit or profit organization doesn’t matter - membership fees are the minumum of 6500 EUR (!!!).

    What do you get for this fee? Getting involved in the discussions on the mailing lists? Anything else? Feel the prestige?:-)

    OK, these are the ways someone can enter the working group as an individual working for an orgaznization or a company…

    You can always try to join the WG as an individual not involved with any organization or company, but I guess it’s a bit harder…

    Hmmm…

    As to the discussions… Sometimes I wonder if it is so important a thing, that we should use B or STRONG or I or EM … and that there are more important things…

    As someone else mentioned, would be nice if the html and CSS standards can move a bit quicker, as the Web itself is moving…

    And, yes, what about Zeldman, Bowman, Meyer, Cederholm? Shall we hear their opinions on that? :-)

    Cheers, Michel / CSS designer, not-liking-any-bureaucracy-but-loving-webstandards-and-things-that-work

  38. May 8, 2007 by zcorpan

    IIRC, Simon Pieters even said once that they were redefining the meaning of ‘paragraph’ to include all sorts of block level containers.

    Yes. See proposed definition of “paragraph”.

    Apparently some people thought it was ‘unsemantic’ to wrap form controls in DIVs, and thought that P elements would be better.

    No, that’s not the rationale AIUI. The rationale is that <p> is used for things that are not strictly paragraphs in the English sense in the real world, so defining <p> to be more generic is a more pragmatic approach than saying that all existing pages that use <p> for things like grouping form controls, poems or postal addresses are non-conforming. In practice markup consumers don’t benefit from trying to distinguish between whether a <p> or <div> is a paragraph in the English sense or something else (what would the benefit be?).

    That said I haven’t said that I agree with the proposed definition — I don’t feel strongly about it either way. If there are arguments against defining <p> to be more generic than a strict interpretation of the dictionary definition then let’s hear it.

  39. I can see the point on that too. There’s always indecision when you have a simple one-line bit of text or instruction that isn’t exactly a paragraph but isn’t anything else either. Do you make up all sorts of other elements to cover all the possible types of text that are not paragraphs or lists or quotes? Or make up a new element that covers all miscellaneous bits of text?

    But, on the other hand, you are still using the paragraph tag to mark up things that aren’t paragraphs which isn’t strictly semantic. If you start to give a little bit on semantics for this tag, then what else comes next?

    I don’t have a strong opinion either but I can see why this is being proposed.

  40. Karl Dubost [W3C Conformance Manager, QA Activity Lead] has an interesting compromise for error handling in Re: Error handling and legacy content

    Parsing model for user agents is a good thing. Draconian handling in some cases might be proved useful too, for example, in authoring tools, CMS and code producing scripting tools.

  41. Matthew Smith:

    Rather than dealing with important issues like how to make HTML more feature-rich so it can be a better rich application platform, we are stuck with arguments on semantics and accessibility.

    So tell me, why exactly is bells and whistles more important than semantics and accessibility?

  42. May 8, 2007 by Roger Johansson (Author comment)

    John:

    What about a petition for us to sign? Listing elements that need saving and points of view that need weight? One person with the support of 100s maybe 1000s could be a more powerful than an uncoordinated group?

    That could be worth looking into. I doubt I will have the time to organise anything like that though.

    Jake:

    The new spec sounds like “If we make theft legal, crime rates will drop”

    Heheh. Yeah, there is a risk of it being interpreted that way, though I’m pretty sure that isn’t the intent.

    mpt:

    Both of these phenomena are demonstrated in the unpopularity of XHTML-served-as-XML, amongst both browser vendors and users.

    I think it’s hard to say much about the popularity of XHTML since the majority browser doesn’t support it. If it had, things might have been different.

    Lachlan: Since we discussed this on IM I’ll just say that I agree with some of what you are saying while I still disagree on some points.

    Tom:

    First of all, it’s laughable - laughable - that they are using a mailing list for this group. How is that effective?

    Actually, I think a mailing list is the best choice of communication in this case. It is quick and immediate, and does not force you to visit a forum repeatedly to keep track of what is happening.

    Matthew:

    Rather than dealing with important issues like how to make HTML more feature-rich so it can be a better rich application platform, we are stuck with arguments on semantics and accessibility.

    Well, I do prefer sorting out any accessibility and semantics issues before adding application-specific stuff.

    Megan:

    It seems like it is possible for independent people to join the group - are there any qualifications to become an “invited expert”?

    Yes, it is possible. I joined as an individual. I don’t know on what grounds applications are accepted or rejected.

    Michel:

    Joining the W3C as a company/organization, is something quite expensive.

    Yes, but it is free for individuals to join the HTML Working Group.

    Sometimes I wonder if it is so important a thing, that we should use B or STRONG or I or EM … and that there are more important things…

    There are definitely more important things.

  43. ref joining the HTML Working Group: http://lists.w3.org/Archives/Public/public-html/

  44. Hi my english isn’t good enough to explain my meanings above semantic , so it doesn’t make sense for me to join an english discussion group.

    And in this special case I’m not so popular like you to become a free place in w3c groups. ;)

    And to pay is too expensive for me.

    My thoughts in “tags”:

    w3c is now an authorithy - but their new way isn’t mine and not the way of so many webdesigner they would like to have a semantic correct html AND an accessibility web.

    And I agree with the 4items in your comment: 19. 2007-05-07, 22.56 by Roger Johansson

    Figurative html5 looks like a a big box for my stockings. And in this big box there are more than thousand other boxes for grey stockings, for grey stockigns with straight seam, another box for grey stockigns without straight seam and another for blue stockings,another for blue stockings with straight seam…….

    I like boxes for my sockets but if they are to many I lost the overview. But I have boxes to get the overview.

    This is the false way to order my stockings and the false way to define html5.

    A Hyper Text Markup Language without semantic can’t be the future. It is the stone age!

    kindly regards Monika

  45. What we probably need to ask is not what we want the web to be now but what we each envision the web to achieve in the future as technologies and innovation carry us further into the realm of computers talking to computers. That’s where the importance of creating a set of standards with semantic value become more important.

    Personally I am very aware this is a decade old concept running on 30 year old technology. We don’t all envision the web will just stay this way forever do we?

    I may be argueably wrong in my opinion but what I’m suggesting is there are two ways to see this one - the first is that the Web will remain about people talking to machines (in which case there’s no foul probably that junk code is used) and the second is that the Web is becoming more about applications and so forth communicating with each other (in which case a clearly defined meaningful communication conduit is imperative).

    I wouldn’t say there MUST be 100% semantic hardcore use of HTML on any site. I would say there must be the future ability in the specification to provide a semantic framework which meets that future potential.

    (or maybe we’ll find an evolution occurs where a new version of all this is superceded by a more effective child which does provide us all of that - the W3C does run the risk of becoming less relevant if it makes the wrong decisions and it must realise that “perception” of W3C is just as important in our acceptance as is their actual technical ability)

    OK rant ended… sorry roger and I admit much of what I said is probably not what everyone needs to hear.

    So in short I pretty much agree with you.

  46. My friend Joe D has been sending me samples from the Working Group discussions. There was a big debate about adding an element called indent, basically because of people not bothering to learn style sheets. Oi veh!

    Vendors and hacks (like back end developers who only understand tables and font tags) are happy with web sites created with the equivalent of Bondo body filter and duct tape. So disappointing.

  47. May 10, 2007 by chris Jangelov

    This discussion started with Rogers warning about standards, semantics and accessibility being at risk to be left out in the HTML 5 working group. In the following diskussion there is also a warning that we mustn’t break the web, i.e. existing tag soup has to be interpreted.

    In my mind a few thoughts emerge:

    1. Tag soup is working today. Otherwise the web would alreday have been broken. So that can hardly be a major reason for developing a new HTML standard.

    2. W3C is an authority today. The European Union refers to W3C as publishers of standards for their i2010 initiative (That is making information accessible to all EU citizens - and that means ALL information, not just governemental information.)

    3. As Steven Clark pointed out - we are not just talking about people talking/listening to machines but also machines communicating with machines. Could we envision web services without a rather well defined set of standards? Could we envision web applications without some kind of strictness in the coding? I’d say no.

    4. I was in a meeting once with people who converted instructions and manuals to xml to make them available to technicians in the field. If you are on a ship there are a multitude of technical systems that has to have service and to be repaired. There is too much knowledge needed for any crew to have total updated information in their heads. So they have to have it available when they are actually doing the work and it has to be easily updated and distributed. Hence the need to put it on-line. When I asked them about the man hours needed to convert written information to on-line information they had a rather interesting example. If their client had used formating templates when writing in MS Word, the job vas done in weeks. If they had used buttons for bold, tabs for indents, and so on… - it was a question of several months since everything har to be rewritten (or copy-pasted). Just figure the price tag for not separating content/semantics from form when the work was done first. If I were a shareholder or member of the board I wouldn’t accept paying twice for the same work. This is not difficult technical knowledge - this is ignorance.

    5. If I as a company pay for putting a lot of information on-line I would not accept having to pay twice because my visitors use hand held terminals/cellphones. This is only a question of marking content for that channel and format it for that media. I would not pay a programmer that had put me in this situation. I would sue him/her for the time delay.

    6. Before writing non accessible code you have to decide exactly who you are going to shut out. The visually challenged? The elderly. The people in bright rooms. a. The visually challenged. Just don’t separate content from form and you get rid of most of them. b. The elderly. Now we include the design profession. Just make long droplists and multi level fly out menus. Problem solved. c. The people working outside or near windows. Time for designers again. Just use gray on off white background. Or shall we say light blue text…? (Sorry for broadening the diskussion to include graphical design, but one point is that we are dealing with different professions. Content creation; Site design; Code; Graphical design (and others). Accept that and you don’t have to cook any more tag soup. The web master IS dead.

    7. The on-line world is (in a way) a mature industry. It is not lack of knowledge that is the problem. It is people with a great lust to send bills and a lack of lust to grow professionally. Or a lust to cheat in other professions.

    And what would be the conclusion of this truth yelling?

    Well..

    A: We need a webb that works for “everyone”.

    B: I don’t want to pay twice.

    C: We need standards and hence an accepted standard issuing body.

  48. Thank you for writing this post, it says exactly what I have been thinking for quite some time now.

  49. Well, I’m an invited expert now…

    but how the heck do you actually participate in a discussion?! There is no help at all in guiding new users in how to do anything. I can see the mails, I can read the mails…I’ve replied to mails - only to not see any of my own mails appear in the archive. What am I missing?

    Good lord what I’d give for a forum instead of this archaic mess littering my inbox. Is it any wonder things are so slow at the W3C? I get the distinct impression mailing lists are used simply because they always have been. I’ve been on the net for 6-7yrs, and I’ve never had cause to use a mailing list. It’s an outdated form of communication that I’m not familiar with. Why are the W3C still stuck in 1999?

  50. May 10, 2007 by mattur

    It’s painful to accept that everything you’ve been told about web standards is wrong. To have one’s worldview (webview?) abruptly displaced by a new, different worldview, championed by people who come from a very different, technical perspective is always going to create friction.

    But remember, the only thing that has changed is that the two groups have swapped roles.

    Those of us who previously disagreed with the W3C have had to put up with W3C-supporters evangelising an approach that made no sense to us.

    That frustration you’re feeling, Roger, that deeply-held conviction that something is going wrong? That’s what we’ve been feeling for the past 10 years.

    Our roles are now reversed, and it’s up to those of us who previously criticised the W3C to put our money where our mouth is and actually deliver improvements to the web for everyone.

  51. May 10, 2007 by Stevie D

    you should focus discussions on how to release a new version of HTML and CSS more then once a decade.

    This a common phenomenon, and W3C are unlikely to have much effect on it, in practical terms.

    As the user base grows, and as the technology matures, users are less likely to upgrade quickly. When there were only 100 people using the internet, they were all pretty techie, and browsers were very primitive, so any new release was snapped up quickly. Now that millions of people are on the internet, many of them unable to do anything more technical than type a URL into Google (it frightens me how often this happens…), they aren’t going to be so quick to upgrade. IE6/7 “works fine”, they don’t see anything wrong with it, so they will carry on without changing anything. The only time they will “upgrade” is when they eventually buy a new computer.

    What has held back website development in the last few years has not been W3C’s sluggish attempts to get new versions of X/HTML and CSS - it has been Microsoft’s failure to provide browsers that work to the current specs. There’s no point in rushing to finalise new specs when 50% of people are using IE<6 and 85% of people are using IE<7; authors won’t be able to use the new features reliably because most people won’t be able to access them.

  52. May 10, 2007 by Roger Johansson (Author comment)

    Matt: FWIW, I think a mailing list is the most appropriate communication tool for the HTML WG. I tend to forget about the forums I am registered on. A mailing list avoids that problem.

    mattur: Umm… care to explain exactly what you are talking about?

  53. May 11, 2007 by John Foliot

    Lachlan Hunt:

    That is not true. We are against semantics for the sake of semantics. In other words, adding a new semantic feature actually has to have:

    • Real use cases to explain why authors would use it. If there’s no reason to use it, authors won’t.
    • Solve real problems. We need to identify the limitations with existing HTML, figure out if existing HTML can already solve the problem; and if not, why not? If existing HTML can’t solve a problem, then we can consider finding a solution.
    • Provide some incentive for authors to use it. A feature in the spec that won’t be used is no better than a feature not included at all. So any proposed feature needs to show that there will be some real incentive for authors to use it.

    Alright Lachlan, let’s take a look at this then:

    1. Numerous use-cases have been suggested on the mailing list that have been ridiculed or dismissed by other members. A common example is indicating why a piece of text has been italicized, and there are many possible reasons. While it may seem obvious when viewed typographically (a termed used more than once in the debate), there is currently no method when visualization of typographic conventions is not available - i.e. for the blind. The discussion has also pointed out that in braille text conversion, this type of granular distinction is available, although apparently this is not enough evidence of “use-case”; as well, no less than IBM has recently posted a paper which states in part, “…We assert that the only way to accomplish this is to encode the semantic information of the page directly into the page.” http://www.research.ibm.com/journal/sj/443/harper.html.
    2. The real problem is apparent to anyone who understands accessibility issues. Elements such as < i > and < em > provide visual rendering on screen that is missing when you can’t see the text. And if you italicize something, why? How can this be a complicated issue to understand? Currently HTML cannot solve this problem, so a new means needs to be added.
    3. Regarding incentive: believe it or not, there are many, many web authors who work for organizations and institutions that either are mandated to, or absolutely desire to ensure that they meet their moral obligations. Increased semantic markup also assists when it comes to machine readable texts - a hugely beneficial aspect to libraries, educational institutions, governments, etc. I do not understand why it should be the role of the WHATWG or the new HTML5 WG to decide that, because there may be no perceived benefit to some kid in his underwear cranking out a web-page in Mom’s basement, important and essential constructs currently missing in HTML should be left out. Who asked you guys to decide? If you don’t want to use the semantic-meaning enhancements, don’t… you don’t need to use the height and width attributes of the image element either…

    Proposing solutions like the role attribute and RDF without clearly identifying what problems it really solves and why those problems can’t be solved with existing markup, doesn’t help. Personally, I don’t think role is a good solution to anything. It’s over engineered and far too poorly defined.

    Since I started the thread Subject:@role, I feel compelled to take that comment on. Personally, I do think that role is a good solution, so my opinion negates your opinion, and we are left at square one: so much for opinions.

    @role has been discussed and worked on for over 2 years now, and it’s intent, it’s capability, and it’s scalability is apparent to many. Once again, no less than the good folks at the Mozilla foundation also think that @role has a use; witness the emergent work with the ARIA suite, and the part that @role will play in future AJAX development - and who’s funding that work? IBM. The WHATWG people want a better web, but apparently if it’s not their idea, it’s dismissed.

    The problem has been articulated in many different ways, by many different voices [Lauke, Holmboe, Korpela, et. al] to apparently deaf ears. The counter proposal, to use the @class attribute to assign semantics has been rigorously picked apart by many accessibility specialists and long-standing members of the W3C, but hey, the WHATWG folks think it’s fine; they see absolutely no problem with: < span class=”copyright” > the fine print in the footer< /span > (geez, i don’t see any copyright info there, do you?) You keep going on about the under-informed web creator who doesn’t get it, and that you just need to make HTML5 simple to use…the problem is, if you dumb it down too far it becomes useless, as this case illustrates - too many Dreamweaver jockeys used the class=”copyright” to mean “…the fine print in the footer” - and the argument that this may break backward compatibility seems to have been dismissed.

    Encouraging authors to do the right thing is a far more effective and less harmful than trying to discourage (or worse, punish) them from doing the wrong thing.

    Give them the proper tools then, and stop asking them to use a screw-driver to drive a nail… it’s amazing what a new set of tools will do for even the reticent worker.

    From my perspective, there is a lot of push of ideologies that are not supported by reason or logic, yet do have reasoned and logical arguments against them.

    … and for many of us, and our perspective, the curent crop of authors that are WHATWG/HTML5 WG is being as deaf and blind to our issues/observations/suggestions as they accused the W3C of once-upon-a-time. Witness:

    It’s painful to accept that everything you’ve been told about web standards is wrong. To have one’s worldview (webview?) abruptly displaced by a new, different worldview, championed by people who come from a very different, technical perspective is always going to create friction.

    But remember, the only thing that has changed is that the two groups have swapped roles.

    Those of us who previously disagreed with the W3C have had to put up with W3C-supporters evangelising an approach that made no sense to us.

    That frustration you’re feeling, Roger, that deeply-held conviction that something is going wrong? That’s what we’ve been feeling for the past 10 years.

    Our roles are now reversed, and it’s up to those of us who previously criticised the W3C to put our money where our mouth is and actually deliver improvements to the web for everyone. Mattur

    There have been many, many reasoned and logical arguments put forth arguing for a better, richer semantic tool set. The arguments we’ve been countered with have all been essentially, “who’s gonna use them”, to which I can only say - add them, and let’s find out… a much better way forward than leaving them out because you don’t have the answer.

    Oh, and mattur, the frustration that we are feeling is that now that the WHATWG was invited back to the table at the W3C by TBL, there is a certain swagger that is coming across that some of us find disconcerting - you guys think you have all the answers: you don’t.

  54. May 11, 2007 by mattur

    …there is a certain swagger that is coming across that some of us find disconcerting - you guys think you have all the answers: you don’t.

    A truly fascinating and deeply ironic comment. Don’t you see? That’s exactly what me and many other people have been feeling about the old “Web Standards” movement all these years(!)

    Do you want me to cite examples? Should I condemn you as a shill for Microsoft, someone who doesn’t want open standards for the web, someone who doesn’t “get it” - simply because you now disagree with the W3C? That’s what it’s been like for the past 7 or so years for anyone who questioned the W3C’s approach.

    Our roles are now reversed: people who advocated the W3C’s old approach are now (rightly) questioning the W3C’s new approach. People like me who previously disagreed with the W3C are now advocating the W3C’s new approach.

    The difference is, anyone who disagrees with the new approach can get involved and argue their case. Previously anyone who disagreed with the W3C was at best ignored and at worst regarded as some sort of heretic.

    Even though I disagree with your POV, it’s great that people are now, finally, questioning the W3C. Long may it continue :-)

  55. May 11, 2007 by Roger Johansson (Author comment)

    mattur: I still do not understand what you are trying to say. Who or what are you representing?

  56. May 12, 2007 by John Foliot

    mattur,

    You clearly have little understanding of standards, of concencus, and of the international ramifications of what is being discussed. This is not some game where you pick a winning side. If you or anyone else who holds a similar opinion think that you are going to “win” something, you are in for a rude surprise.

    Large corporations, governments and educational institutions fund a large part of the work done by the W3C, and if you think that this kind of half-baked poppy-cock will get through to final specification without addressing the very real concerns being expressed by accessibility and standards advocates then you are living in some kind of dream world. Thoughts of “turning the tables” simply prove how immature some of the arguments are that the accessibility and standards advocates are getting.

    Ensuring and improving on-line accessibility is entrenched in law in most if not all of the “Western World”, and any advancement of HTML that backs away from this, or ignores it, will die on the vine. To date, the advocates have been arguing merit, but if it really does come to push and shove, the legal trump card is clearly in our hand.

    Wake up and smell the coffee my friend.

    (Roger, apologies if this heads into the “name-calling” arena, I’ll stop here, but I mean, really???)

    JF

  57. May 12, 2007 by mattur

    I don’t represent anyone but myself, apologies for any confusion. There is more than one way of developing standards for the web. One way is to see web standards as a way of enforcing a “tidying up” of the web:

    • remove presentational markup
    • aim for pure semantic markup
    • make the rules stricter
    • this is the way we should have built the web in the first place

    That’s roughly the old W3C approach, evangelised by the “Web Standards” movement and accepted by many as self-evidently the right thing to do. An alternative is to see web standards solely as a way of standardising the web:

    • standardise presentational markup where necessary
    • semantic markup is a means to an end, not an end in itself
    • keep and clarify the current, undeniably successful, rules
    • the way the web was built was a fundamental to its success

    and my personal favourite:

    • add new stuff that we’ve been waiting for since 1995

    That’s roughly the WHAT-WG approach, now adopted by the W3C. TBL’s infamous Oct 2006 announcement wasn’t just about syntax, it was about changing the W3C’s whole approach to HTML markup.

    People who believe in the old W3C approach are inevitably going to criticise HTML5 because it takes an entirely different approach to the one they’ve come to associate with web standards. They’re still looking at HTML5 through XHTML2-shaped glasses.

  58. May 12, 2007 by Roger Johansson (Author comment)

    mattur: Thanks for clarifying. For a while there I thought you were an advocate for the Group of Misnested Table-tag Soup Users ;-).

    As you probably realise I am more for the “old W3C approach”, but I don’t think XHTML 2 is the way to go. In fact, I don’t feel that I strongly disagree with any of your points, neither the “old W3C” ones nor the WHAT-WG ones. “Standardise presentational markup where necessary” does seem useless to me, but could possibly be done without causing too much harm.

    No XHTML2 or WHAT-WG-shaped glasses here :-).

  59. May 14, 2007 by mattur

    As you can probably tell, I’ve got HTML5-shaped eyes… :-)

  60. Sounds pretty bad.

    I am currently also planning to join W3C and propose some semantics, which I am thinking about and which might be useful for the future of the Web.

    I am fully with you regarding this matter.

    Nice looking blog as well BTW. Well done and keep it up.

  61. May 15, 2007 by John Foliot

    and my personal favourite:

    • add new stuff that we’ve been waiting for since 1995

    Allrighty then, consider this: web accessibility advocates have been waiting roughly the same amount of time for web developers to understand that there is more here than just pretty pictures and design layout. The ability to mark-up content with structure and logical intent has opened the world to many, many people with various disabilities. I’ve seen and worked with blind users who remark how much the web has given them independence; I have a friend and associate who is completely paralyzed - the web is his lifeline to the world. I’ve seen work done that uses the web to allow the most severe Downs Syndrome patients communicate with each other and the outside world.

    So that “waiting for new stuff” since 1995, for me, and many others, includes extending the already wonderful communication outlet that is the Internet to members of our society that are often marginalized.

    You ain’t the only one waiting mattur…

Comments are disabled for this post (read why), but if you have spotted an error or have additional info that you think should be in this post, feel free to contact me.