@media 2007 Hot Topics Panel

• Listen to the audio recording of this panel

Jeremy Keith: Good afternoon, everybody, and welcome to the Hot Topics Panel. [sizzling sound]

[laughter]

Jeremy: This was a lot of fun last year. Hands up who was here last year. Okay. I think you’ll agree, we had a lot of fun, because Joe wasn’t there.

[laughter]

Jeremy: No, we had a lot of fun with the Hot Topics Panel, and we’re going to have fun again this year, with some questions provided by you, the audience. We’ll try and make this a bit more interactive. And I’ve got a lovely lineup of people here.

On the far left of me over there, we have Mr. Drew McLellan, blogs at allinthehead.com. He’s co-lead of The Web Standards Project, along with Kimberly Blessing, which in a weird sort of way sort of makes him my boss, in some sense. And he now works at Yahoo!, where he takes part in many meetings, lovely, lovely meetings.

[laughter]

Jeremy: Over at the other side, we have Mr. Richard Ishida. Am I pronouncing that correctly? “Isheedah,” or “Ishiduh”?

Richard Ishida: “Ishiduh.”

Jeremy: Mr. Richard Ishida from the W3C, who knows all about many cultures and localization and internationalization, and he’s a very smart man. He’s come all the way from Harpenden to be here today.

[laughter]

Jeremy: On this side of me, I have Mr. Joe Clark, the man who described himself once as “the only X in the village, regardless of what X is or where the village is located.”

[laughter]

Jeremy: Which is a pretty good description of Joe. I think he would describe himself as a resident of Leslieville, Toronto, but he’s not really; he’s actually from Spud Island in Canada.

And on this side of me, I have Mr. Dan Cederholm, who you probably know from his blog, simplebits.com. He makes me insanely jealous with his design props. And I can now do something I’ve wanted to do with Dan for quite a while, which is, I owe Dan a bottle of wine…

Dan Cederholm: [laughs]

Jeremy: …because he very graciously allowed me to use this term, “bulletproof” for my book, which I ripped off completely from his fantastic book, Bulletproof Web Design. So I just went out and picked this up. It’s a bottle of Pinot Noir from Chile.

Dan: Can we open it right now? [laughs]

Jeremy: Yeah, let’s have some.

[laughter]

Dan: I think this panel just got a lot better.

Jeremy: Now, let’s kick off. Down to business.

Remarks about the all-male lineup have been stricken from the audio record of the Hot Topics Panel

Jeremy: Ah! Here’s an interesting one: “Is the jump to CSS3 too big, complicated, and taking too long? CSS 2.2 anyone?” Now, what this is referring to is an idea by my esteemed colleague, Mr. Andy Budd, if you’ll make yourself known…

There he is, Mr. Andy Budd, down in the front. He floated this idea at the Highland Fling conference a couple of weeks ago, a couple of months ago, up in Edinburgh. The idea being that CSS3, it is taking a long time, and there are good reasons why it’s taking a long time. There’s some very complex stuff in CSS3. I don’t know if you all are keeping track of what’s going to be in it. A lot of it involves text and internationalization issues, but there’s some stuff that will be ready to go: multiple background images, things like that. Why don’t we just release that stuff now and worry about the hard stuff later, incrementally?

So maybe I will turn to our representative from the W3C on this panel, Richard. Is this a reasonable plan, CSS 2.2?

Richard: Well, I think it’s an interesting plan. And I think it’s a plan that people should propose to the W3C, and I think you will find that they are interested in your views. Certainly, yeah, it has been taking a long time, and I’m as frustrated as anybody else. Because CSS3 contains a lot of, a fantastic amount of internationalization features, to support typography around the world.

The problem that we have, really, is just resources to get the work done. We’re still working on CSS 2.1, trying to get it out of the door. And everyone at the W3C wants to get it out of the door as quickly as possible. We need help. It’s complicated stuff. In some cases, like the vertical text that we’re working on, for example, and line breaking in Japanese and Thai and so on, we need some fairly exclusive resources, some people with special knowledge.

But the W3C is a place where we welcome people to help us to develop the technologies for the Web. We don’t have a monopoly on development of technologies there, we don’t say, “It’s just us that now have to do it,” at all. We would definitely welcome more people getting involved, and that would help move the work forward faster.

The idea of splitting it up into smaller segments is a really interesting one in my mind, but, again, please don’t just moan about it, or talk about it in designer circles; come to the W3C and say, “Hey, here’s an idea, would this work?” and I think you’d find that there’d be interest in discussing it.

Jeremy: Drew, do you think this is something the WaSP could coordinate, as in trying to gather people together to put together test cases, or ideas for how modules might work?

Drew: Possibly; what the Web Standards Project aims to do is encourage people to use the standards that are there, and to encourage makers of tools to implement those standards. We haven’t had a particularly active role in helping define those standards, up to this point. I’d be interested to find out whether input from browser manufacturers, in the form of implementations, is a welcome way to introduce ideas. You said that Andy floated the idea; I thought it was actually quite a well-positioned argument.

Jeremy: Okay.

Joe: And didn’t we have one of the co-editors of the CSS spec, Mr. Håkon Lie, here today, saying it was a good idea?

Jeremy: Yes, yes we did.

Joe: So pigeonhole him right now—

Jeremy: He’s left the building.

Joe: Oh, bugger.

[laughter]

Jeremy: Dan, can I ask you if you have a wish list of things that you’d like to see right now?

Dan: Right now, today?

Jeremy: Yeah.

Dan: Multiple background images, for sure. Right? Who’s with me? [laughs]

[applause]

Dan: That would alleviate a lot of the problems we have, as designers, making things interesting.

I would love to see CSS 2.2. My only question would be, “Would that speed up the process in terms of the browsers implementing it, or is it still going to take a long time for browsers to support that stuff?”

Jeremy: Yeah, we’re going to be in an interesting position now, because we’re not waiting five years for Internet Explorer to release a new version. They’re actually going to be releasing stuff pretty rapidly—

Dan: That’s true.

Jeremy: So maybe the specs should be released at an equally rapid pace.

Following off this question—this is one that maybe I’ll turn to Joe for—is this: “With the WHATWG and the criticism of WAI, is the W3C as important to modern Web design as it once was?” Joe?

Joe: No. I mean, demonstrably not. They’ve succeeded in doing a couple of clever things, though. First of all, they’ve assimilated, in a Borg-like fashion, the WHAT Working Group; and the WHAT Working Group are turning out to be the same kind of assholes that the W3C were.

[laughter]

Joe: I can has summary attribute, right? Nonetheless, they’ve squelched that. They essentially bought the company and killed it. It’s a typical Silicon Valley approach, and that’s fine.

Now, to their credit, the Web Content Accessibility Guidelines Working Group read and responded to absolutely every objection to the first draft of WCAG 2. Every single one, they didn’t skip any, and they tried to do something to address all of them. While I haven’t read the current draft of WCAG 2, I read the two changes documents, which show that there has been tremendous improvement. So I’m hoping to propose an article to Zeldman about the current version of WCAG 2, which will probably not begin with the phrase “to hell with.”

So, unbeknownst to me, even though I follow the lists and so on, they had essentially taken everyone seriously, even me. So I think that the flipside of all that is, it seems like they’re trying to co-opt the competition. But almost everyone in this room can write a standards-compliant Web document, and an accessible Web document, without ever checking a spec. We’re in a post-guideline age, a post-checkpoint age, right? And that will continue, because there will be more and more standardistas and fewer and fewer people making sites with 37 tables. So in that respect, there’s, I think, going to be the dominant one in the future. The checkpoint, guideline, specification process, blah blah, is going to wane.

Jeremy: Richard, as the representative of the W3C, to address those criticisms, the W3C getting into bed with WHATWG—and yes, things are getting more agile—but it’s 2007 and there is serious talk about the font element being in HTML 5.

[audience member boos]

Jeremy: Thank you, Andy.

Dan: [laughs]

Jeremy: What’s going on? Surely the W3C is completely irrelevant when there’s serious talk about having the font element, in this day and age.

Richard: Well, the W3C doesn’t have a monopoly on good ideas. And we are not the only people involved in standardizing the Web. I mean, JavaScript is not a W3C thing. We have the IETF, and there’s other standards organizations out there as well.

So I think that what we have is a base set of technologies that we’ve been working on for a long time, and which are working quite well. And hey, you have HTML. It’s thanks to the W3C. And we have taken on board the things that you’re saying, that we need to develop HTML as a language further, and we’re working on that.

I think, though, that what’s important for me is, we reached this point recently where there was quite a lot of criticism of the W3C. And for me, the object of the criticism wasn’t quite right. I think the problem we have at the moment is communication rather than anything else. And the way that I see it, there’s, in particular, three main groups. There’s users, there’s browser developers and tools developers in general, and then there’s standard organizations.

And what I would like to see more of is communication between the users and the W3C and vice versa, but also communication between the W3C and the browser developers. Because we talked about CSS, and we’ve had complaints at the W3C that CSS specification is going too fast, because the browsers can’t keep up with them.

But I’d also like to see communication between the users and the browser developers or the tools developers. And I’m including their editing tools as well as browsers, because those are fundamentally important to achieving the effects that you want, too.

And there has to be this sort of triangular flow of communication going on. And I think that that’s actually the problem at the moment. And if we can get that flow of communication going, in a supportive way rather than a critical way, saying, “We need this, guys, and let’s take on board your ideas,” I think that that’s the key.

Joe: How are you defining users there? Do you mean, by users, all the people in this room who are developers or designers, or do you mean my friend who just uses the Web to surf for porn?

[laughter]

Richard: Well, I mean all kinds of users. And obviously, the W3C can’t have a telecom with the whole world, but we’re trying to reach out to people in whatever possible way we can. And we’re hoping that people in Thailand who can’t get Firefox to wrap their lines properly will talk to the people at Mozilla, and so on and so forth. But certainly the people in this room as well, yes.

Jeremy: Drew, do you see the Web Standards Project as doing exactly what Richard says and getting directly in touch with people like browser vendors, tool vendors, and almost bypassing the W3C, completely going from us developers straight to browser vendors, without going through the standards body?

Drew: Well, that’s definitely the danger. And it’s a problem that we saw back with the version three and version four browsers, where the browser manufacturers were just taking it upon themselves to innovate and to leap ahead, regardless of any specification. Whereas that created a lot of problems, it also gave us a lot of interesting innovation, and things that have solidified and specs that have been built up, based on those implementations.

So I think that we’re in, I say in danger, but the situation may occur that the browser manufacturers are then just pushing ahead, regardless of any spec, because the specs aren’t there. And then we’ll just see what we end up with; hopefully not too much of a mess. Perhaps we’ve had our fingers burnt once, and now we’re better.

Jeremy: The de facto Web standards project, wouldn’t it be?

Drew: Maybe.

Jeremy: Dan, do you care about W3C?

Drew: [laughs]

Dan: What are we talking about again?

Jeremy: Yeah.

Dan: [laughs] No, I’m just kidding. I do, I do.

Jeremy: I mean, have you signed up to, say, the HTML 5 mailing list or anything like that?

Dan: I haven’t, actually.

Jeremy: You haven’t? okay, good.

Dan: I’ve sort of stepped back a little and just wait to see what happens; which, maybe I should be more involved. Maybe that’s a bad position. Like Richard was saying, we need to get more people involved. When I hear something like the font element coming back in HTML 5, it scares the hell out of me. Up to this point, I have been sort of putting the blinders on, but I think I need to get back involved.

Jeremy: Yeah, take a look at what’s going on.

Dan: Yeah.

Jeremy: Now, I believe we have one more related question on this topic, then hopefully we can move on. This is about WHATWG again. “Ian Hickson,” that’s Hixie to his friends, “recently wrote about a need for a WHATWG style effort with CSS. Do we need this?”

So I guess the same kind of grassroots, agile movement from the people who brought you the font element in HTML 5, now they want to look at CSS. Do we think this is required? This is kind of following on from the previous question, I guess. Do we want the grassroots approach to CSS, as opposed to a specification approach?

Drew: Well, perhaps if they were to finish developing HTML 5 first, see where we are, then they can start messing with something else. Because if you look at what they’re doing with the… You asked Dan if he subscribed to the mailing list. I’ve been subscribed to the mailing list, and they’re just really dull.

It’s not an easy thing to follow. It’s a fairly technical, specialized subject, and it’s a lot easier to say we need people to get involved than it is actually to get involved, because it’s quite a serious skill set you need, to be involved in an active way.

Jeremy: Yeah. My impression with CSS3 is that the issue is not, “Well, what are we going to do? We need ideas.” It’s more implementation and all these problems that Richard pointed out. Would that be fair? Where do we start, is more the question.

Richard: Well, I think ideas are important, too. This goes back to what we said earlier. If you guys out there know that, what was it, multiple backgrounds are something you really, really, really need, let us know.

What worries me about the WHAT Working Group going off and doing something independently is whether we will come back together again at the end. And I don’t think it’s in anybody’s interest… we’ve finally got the standards right and rolling, and we understand the value of having one way of doing things. I mean, I can remember trying to write Web pages 10 years ago. And with JavaScript, every line you wrote was a pain in the neck, because there was no standard way of doing things, and so on and so forth. And I really don’t want to start going back down that track in the wrong direction.

I look forward to getting more user feedback. I look forward to getting the user community telling us what we need to be working on, and I think the whole of the W3C wants that as well. I worry a little bit about fragmenting standards.

Jeremy: Okay. But generally, the advice is—if somebody wants to bitch about something that they want to see implemented soon, they want to see it—bitch loudly on a blog, and show examples of what you want and how you imagine it working.

Richard: Yeah.

Jeremy: If you’re going to bitch, bitch properly.

Richard: And bitch nicely.

Joe: We’ve been doing that for years and they ignore it. They say you must submit something to the list, and you have to be a subscriber to the list to do that, and to be a subscriber, you have to be an invited expert. Now, we’ve tried this. The multiple background images thing is the classic example. We’ve been asking for that for five years. Where the hell is it?

Anyway, the problem with CSS, of course, is the permutations, right? I have written valid sets of nested list elements that required Zeldman to rewrite his CSS, okay? I have stuff on my personal Weblog that Opera can’t display properly, even though it’s just UL, DL, OL, LI, DT, whatever the hell it is, right?

Jeremy: LOL

Joe: So yeah, this sort of thing. LOL, whatever the hell it is. So the permutations are astronomical.

So if someone came along and, say, did a nightly build of WebKit or a nightly build of Firefox with “brand new CSS properties,” well then, for that even to be remotely viable, they’d have to test it against… Say, if they just, for example, take some of my multiple indented list markup and put their stuff in it and see if it works, which is a good test case.

So they’re not comparable. HTML and CSS are not comparable, because the permutations are so much bigger in CSS because of the cascade.

Jeremy: Still, I don’t think you can have too many test cases. So, published test cases, I think, are good. The W3C does pay attention to blogs. And so do browser vendors. I know that the Microsoft team, especially, are scouring blogs. They do pay attention. I know that the Opera people pay attention. Your voice is heard, is the message.

Joe: And we have David Storey from Opera. He’s an excellent person to send things to, actually.

Jeremy: Good plug for David, then. Moving on, maybe we can finally get away from all of this W3C talk.

Oh, I like this. “Which site would you most like to redesign, and why?” This isn’t just for designers, obviously. This is for anybody with frustrations. But I will turn it over to Dan first, I think.

Dan: What site would I most like to redesign? Actually, this is going to be a local one, but Boston.com. My home city. I use the site every day, and it’s not particularly designed well. I won’t go into the details because it’s not going to be relevant for anybody unless they’re from Boston, but that’s the one.

For instance, I’ll give one example. There’s a… You go to the home page. There’s a panel with three tabs. Usually, it’s like news, sports, and weather. I’m a huge Red Sox fan, so I often hit the sports tab to see the latest sports news. But, it’s a panel that’s driven by JavaScript, so the panel changes dynamically. I’ll go to a story, a sports story, and then hitting the back button to get to another sports story resets the tab set, and I can’t find the story again unless I go back to the sports tab.

So, that was kind of a long story, wasn’t it? Yeah, Boston.com.

Jeremy: Okay. And it wouldn’t be as simple as slapping on a user style sheet to make it look nice?

Dan: No, but I wonder if I could solve that particular problem with a user style sheet. But, in general, the rest of the site is kind of crappy, too.

Jeremy: Richard?

Richard: So, I want to say my own, but I guess you want me to say the W3C site.

[laughter]

Jeremy: It was a leading question.

Richard: Don’t worry, we’re working on it. Well, you might still want to worry, but we are looking at improving the W3C site.

Jeremy: Would that be the one? You say that’s the one you’d like to redesign?

Richard: Yeah, I guess that’s it. Anyone who would like to help me redesign my own site, I’d be happy to.

Jeremy: Okay, Drew?

Drew: Well, I’d actually like W3C to stop redesigning their site. Every time I go and look something up in the spec, the initial page into the HTML 4 spec has been rearranged, and it takes me several moments to find where I’m going.

Joe: Moments you’ll never get back.

Drew: That’s right.

[laughter]

Drew: Though, if you think the W3C moves slow, something that moves an awful lot slower is the world of banking. The thing that I’d really like to redesign is my online banking. I want to be able to tag my transactions in and out, and then do some reporting on them, and then separate the income and outcome and things into different bits. It’s just like really, really dumb application. It shows you what went it, what went out, and that’s it.

Jeremy: Joe, do I even need to ask?

Joe: No, I have a serious response, which is I’d like any airline or public utility site to be good. Pick any airline, any public utility like the phone company, cable TV, anything and have them actually work. I mean, trying to book a ticket on Air Canada now requires Firefox on a Mac. That’s literally true.

My heartfelt example is a Canadian one, inevitably. It’s a somewhat small supplier of ethnic foods in Canada. Mister Gouda’s. As in Gouda, G-O-U-D-A. Mister Gouda’s. He buys a stake in all the factories that make his foods so he can control the quality. He has 30 different kinds of rice. He has 12 kinds of basmati rice, right? And this is right up my alley. This is the kind of things I eat, right? I would kill to do Mr. Gouda’s because it would be in 19 different languages and you get to the page on rambutans, and it would be in Vietnamese and Thai and English and French.

That’s the one I’d really like to do.

Jeremy: What about the Toronto Transit Authority?

Joe: Oh, the TTC, the Toronto Transit Commission. No, I’ve proposed to them that I merely vet the applicants for the redesign of that — the worst Website in Canada — for Web standards and accessibility. They won’t hire me straight up and I don’t want to go in with some other contractor because we might lose, and I would just not get paid.

Jeremy: Okay so…

Joe: There you go. You had to ask.

Jeremy: I did. I thought that was going to be it.

Joe: Sorry. Mr. Gouda’s.

Jeremy: Mr. Gouda’s. Basmati rice wins.

Here’s an interesting one. I wonder who can take this. “How does the development of the document based Web differ from the development of the application based Web?”

This is begging the question because this is leading us to believe that there are two different World Wide Webs — one for documents and one for applications. As I was saying earlier in my talk, I see it more as a sliding scale. As we get further towards the applications the question of why you’re making them with Web technologies becomes bigger and bigger. It gets a lot harder. The development varies a lot I think.

Drew, you work at Yahoo! You’re starting to head pretty far up that scale towards applications.

Drew: Yahoo! Does all sorts of different things.

Jeremy: Moving away…

Drew: I should point out that I didn’t speak for them. What was the question again?

Jeremy: How does development of the document based Yahoo! Differ from the development of the application based Yahoo!?

[laughter]

Drew: One of the really big things, because traditionally in document based, information based sites, one of the big processes in the information architecture is site config — working out just how everything joins together. That becomes really complex in applications because of a sudden you got a document based application.

You’re having to build functionality with documents only they don’t fit together in that same way. It becomes very difficult in a planning phase to represent how they do fit together. Especially to put that in a document the four levels of manager are going to read and try and understand.

That’s one of the big challenges.

Jeremy: The planning stage I think can be quite difficult with applications as opposed to documents. It’s something I actually want to ask Jason after his talk because he was mostly talking about document based things, articles. When it comes to applications using a two dimensional tool like Photoshop starts to hit its limits. Dan, have you run into this too in designing things that are application elements? Odeo had a lot of dynamic little things going on there.

Dan: Yeah, it’s usually more work for me which can be good or bad. It doesn’t affect me too much other than I tend to code a lot more than working in Photoshop for everything. I’m not Photoshopping form elements or that kind of thing. It doesn’t affect me personally too much. I have enjoyed playing with Ruby on Rails, for instance. That changes the way I design. If I’m able to actually get my hands in the application itself. Rather than just handing off templates. It’s a little more satisfying work. It doesn’t affect me all that much.

Jeremy: Okay, interesting. Richard, any thoughts on this move from documents to applications?

Richard: Not a lot except to say that it’s interesting to see. I used to talk to software engineers about user interface design, and I see us moving back into that sort of area and that my former consultancy is becoming relevant.

So those of you that attended my talk yesterday will have heard about composite messages and how that can really cause you problems when you localize, if you didn’t think about it beforehand. Text expansion and how you don’t really know what is going to come out of the database, so you have to think a little more in advance now, particularly in an international world, about what the layouts going to behave like and so on and so forth.

Jeremy: Something I should point out that while people accuse the W3C of moving slowly, one area where they’re definitely moving quite fast is in the Web applications area, WAI, Web Applications Initiative. No wait, there’s ARIA which is Accessible Rich Internet Applications and there’s some other one. And they’re moving pretty fast there’s some good movement there. Yeah, acronyms.

Joe: It’s Dean Jackson and Anne van Kesteren there.

Jeremy: Yes, they’re documenting XHR.

Joe: They’re the incredibly tall heads of that project.

Jeremy: Joe, do you get worried about this transition we’re starting to see?

Joe: Well, one of my disabilities is I can’t code a lick of JavaScript. So I can barely use Gmail kind of thing, and I have the same frustrations with Flickr that you do, so I scarcely ever think about this. So strangely enough, I really have no informed opinion.

Jeremy: [gasps] Quote that!

Joe: Yeah, really.

[laughter]

Joe: On my epitaph.

Jeremy: Gentlemen of the panel, “What’s your favorite cheese?” Richard.

[laughter]

Richard: Cheese?

Jeremy: Cheese. Jon Hicks must stay silent during this.

Richard: Camembert.

Jeremy: Camembert?

Richard: Nice and soft.

Jeremy: [speaks with French accent] Tres bon. Ah, Normandy.

Yeah, a good Camembert? Jon? Yeah? Good Camembert?

Jon Hicks: Yeah.

Jeremy: Okay. I just want to get your reaction to each answer.

Joe: [laughs]

Jeremy: Dan?

Dan: I’ll go safe and say Gouda, but Humboldt Fog is one that I’ve wanted to try. [laughing] I know Jon hates their proof.

Jeremy: Humboldt Fog, okay.

Dan: Yeah, Humboldt Fog, I believe.

Jeremy: So the Dutch people will be pleased with the representation of Gouda there. Ah, now…

Joe: Either Danish Bimbo or Venezuelan Beaver Cheese. [laughs]

[laughter, light applause]

Jeremy: I should point out that Joe is vegan.

Drew: Normally, I’ll go for a soft cheese, but at the moment, I’m enjoying the Lincolnshire Poacher.

Jeremy: Lincolnshire Poacher? That sounds very nice indeed.

I’m a Roquefort man. Basically a very traditional, good blue cheese.

All right.

Joe: I’m actually starting to miss cheese. Historically, the only thing I’ve missed, as a strict vegetarian, is yogurt. I’m starting to miss cheese more. Can you believe it?

Jeremy: Oh, you’ve got to try Stinking Bishop.

[laughter]

Joe: I may have to sit next to him on the plane.

Jeremy: If anybody is staying in London for the next few days, you should get down to Neal’s Yard in Covent Garden. Simon Willison will give you a tour.

Joe: [laughs]

Drew: Other cheese shops are available.

Jeremy: All right, this is more a personal question. This is a question that we asked in San Francisco as well, but I think it’s a good one. The question is: “What article, presentation, or book has been especially inspiring or motivating to you as a Web professional?” Drew?

Drew: [laughs] I thought you’d come to me first, since I’ve got absolutely no idea. [laughs]

Jeremy: Think, think, think. Okay, you know what? I’m not going to restrict this to article presentation, but just something that’s been especially inspiring.

Drew: I think it’s got to be going back six, seven years, to when Zeldman redesigned A List Apart with CSS. I think that was probably the biggest turning point for me. I was building everything with tables, as we all were, and sort of then pretty much just dropped that and went with CSS. And it was bumpy for a couple of years, but well worth the effort.

Jeremy: Yeah.

Drew: Probably for a lot of people working in the Web at the time, that was a fairly fundamental turning point.

Jeremy: Anyone else have that kind of road to Damascus feeling with Zeldman’s redesign of A List Apart. I know it was pretty big. Doug Bowman’s Wired redesign was quite a turning point as well, I think.

Drew: I think so, yeah. Not forgetting Mr. Cederholm.

Jeremy: Fast Company, yes.

Dan: Oh, right.

Jeremy: What influenced you? And again, we don’t have to restrict this to article presentation, although, if you can…

Dan: Yeah, article, The Dao of Web Design, by John Allsopp.

Jeremy: Oh! I was going to say that!

Dan: Oh, sorry.

Jeremy: No. Yeah, I agree, 100 percent.

Dan: How many people have read that, The Dao of Web Design? Yeah, okay.

Joe: If you haven’t, look it up.

Dan: Yeah, read it. It’s great.

Jeremy: It was written five, six years ago…

Dan: Yeah.

Jeremy: And it is more relevant than ever. The Dao of Web Design, by John Allsopp, on A List Apart. I heartily concur with you there.

Dan: Cheers.

[laughter]

Jeremy: Richard.

Richard: Well, there was a book by Jeremy Keith…

Jeremy: Oh, Pshaw!

Richard: That’s not the one?

Jeremy: No, really.

Richard: No, really?

[laughter]

Richard: There’s a whole bunch. Zeldman’s book and your books are great, actually—surprisingly enough.

[laughter]

Dan: I agree. [laughs]

Jeremy: The one that came to my mind, actually, was Tim Berners-Lee’s book. And the reason was I can remember, at the W3C, just like you guys, we’re working all the hours that God sends and feeling guilty we’re not with our family, and when we’re with our family, feeling guilty we’re not working, and so on. And occasionally you get into this kind of rut, where you’ve got deadlines and code and this sort of thing going through your head. And I sat down and read his book one time, and suddenly, it lifted me completely out of that, and you looked at the whole Web and you said, “What is this all about?” It’s about people communicating around the world, putting people in touch with each other, and enabling people, enabling democracy and stuff like that. It was fantastic, actually.

Jeremy: So Weaving the Web by Tim Berners-Lee.

Richard: Weaving the Web by Tim Berners-Lee.

Jeremy: I agree. It’s a great book, and it does give you this wonderful historical perspective on the Web, because we can be very close to the coal face. It really makes you see the Web in historical context, as part of human development.

Richard: And realize what it was like before we had the Web.

Jeremy: Yeah, completely. Joe.

Joe: Do you want me to give the same overly sincere answer from San Francisco? I had an outpouring of sincerity in San Francisco.

Jeremy: And it wasn’t recorded.

Joe: Several questions. And I can’t remanufacture it. Okay. But no, the punchline is, every morning, I have my coffee, at an undisclosed location, and I always read something. And several dozen times a year, what I’m reading is a research paper, so I read a great many research papers on a lot of topics that are of interest to me; none of which would surprise you.

The bibliography on my research project, openandclosed.org/biblio, or bibliography or any misspelling of it, will all get you to the bibliography. And it’s like 300 items, and I’ve read all of them.

So the big thing is I actually read the research, because—you stole my phrase; it’s not quite regency effect, because San Francisco was two weeks ago—but I have had a sort of Damascene conversion to empiricism and testing things. So that’s the things that are influencing me now, which is hardcore research.

Jeremy: Okay. Well, it sounds like a nice way to spend your time.

Joe: It’s not as sincere as it was in San Francisco. You had to be there.

Jeremy: You had to be there. Honestly, you had to be there.

Joe: You totally had to.

Jeremy: Honestly, you had to be there.

Okay. This is a big question. “Are educational institutions a lost cause when it comes to teaching Web standards?” Ooh. Here’s the follow-on. This is two questions. “Should we expect everyone to be self-taught?” Now, I know it sometimes seems like educational institutions can appear to be a lost cause. I really hope they aren’t. But do we put our faith in them? Drew, do you want to talk about the Edu Task Force?

Drew: A little bit. It’s really difficult, the subject of education, because if you think how the education system works, traditionally, is, if you’ve got to prepare a course, if you plan for a course together, you get that course accepted by whatever the institution is, they then advertise it. People then subscribe to it, people work on that course for three years, and then they graduate. So by that point, the material might be four, five years old.

It’s difficult to think of a way, within that system, a course can be prepared that can possibly be up to date. It’s difficult just to point the finger and say, “You’re doing it all wrong,” because I can’t really suggest how they could do it better, within the structure.

We have an Educational Task Force at the Web Standards Project. And they’re looking at these issues, working with a lot of people in education, and just trying to get people in education up to speed with what modern practices are, and providing resources and just trying to help the thing along. It really is a difficult one.

Jeremy: It’s tough, yeah.

Drew: But I think educating the educators is definitely a good start.

Jeremy: Joe, the second part of that question, “Should we expect everyone to be self-taught?” Do we want to take the attitude of, “Well, I pulled myself up by my bootstraps and taught myself everything I know”?

Joe: It’s a bit masochistic. In the late ’90s, I couldn’t figure out why my HTML 2.0 pages weren’t validating. I was still trying to validate them, believe it or not. I couldn’t understand that you couldn’t put a P inside a P, right? So everyone has a learning curve.

No, there should be instruction. But the problem is, because Web development is not a high class occupation, the big name universities do not really offer courses in it, right? Or perhaps it’s tacked on in your art degree as a bit of interaction design, which is Flash programming or Fireworks - I don’t know what the hell it is. And then Canada’s even worse again, because, here we go again, the Canadian English, we have a distinction between colleges and universities.

Colleges are essentially community colleges, where you go if you can’t really get into a university and all the Web design training is at the college level. There are a zillion of these across the country. They’re all penny-ante two-bit institutions who don’t pay their instructors well and work them to the bone. You get like four grand for a two semester course, right? That includes all the marketing and other things you have to do.

So you are choosing the runt of the litter at that point. So not only are these institutions atomized across my country, they don’t have the budgets, they don’t have the prestige, and so they get crap instructors.

These are exactly the wrong kind of people to walk up to and say, “We stopped doing tables five years ago.” Because they won’t know any alternative and they’ll get defensive.

Jeremy: Okay, but there are also people in third level institutions and secondary level institutions, who are doing a lot of really good work, and it’s a lot of uphill work and we do need to recognize the good work that goes on inside these institutions.

Joe: Well, honestly, I do know someone back home who is a JavaScript instructrix — she’s a woman — and has a savant-like ability to determine when the kids were cheating off of each other just by noticing the patterns in everyone’s work and so on. And she only taught correct things according to the Web standards. So bravo for her, but that was the only one person in one of the campuses of one of the colleges in one city.

Jeremy: Okay. They are out there. And I’d like to give a special shout out today to Mr. Matt Harris — make yourself known Matt. There’s Mr. Matt Harris. He’s faced an uphill battle in Winchester, the school there. They are out there. There is David Smith here in London, working in St. Paul’s School for Boys. You need to talk to David Smith. I need to put you guys together. So they’re out there. Okay, you all need to talk in to each other.

Joe: Okay, look up emmajane.net — that’s the woman I was talking about, Emma Jane Hogbin.

Jeremy: Okay, same conversation as last year. Yeah, it’s true. Good enough, but I think, maybe together we can do more than we can separate. It’s an uphill battle. Richard, do you see WC3 as the endpoint for education or do you want to liaise with educational bodies?

Richard: I don’t think we can educate the world. We have enough trouble writing standards and things like that. What occurred to me was sort of a world-wide-web type approach to addressing this problem. I was thinking, some of us as developers hold meetings where we get together and we talk about development strategies and stuff like that. In some of these meetings perhaps you could invite people in your local area, who educate kids, to come along and you teach them. Obviously, we don’t expect them to get to the same level as you are.

I mean, I am constantly struggling to keep on the bottom rung of this myself but we could give them some basic education as to why standards are useful and whether they should use XHTML or not, and so on and so forth, and separating out CSS and stuff like that. So sort of grassroots type things like that, might be something… If Patrick wanted to organize a conference for educators I would volunteer to come and talk at that, and maybe some of the people would. Pardon? You don’t think they would come?

[inaudible]

Jeremy: Actually we have microphones. If you could just hold on a second, we could get one of the oompa loompas to send you a mic. Run, run from me, oompa loompa!

[laughter]

Go on, run!

Woman: They won’t come. In Bristol we ran skills course, interestingly which we took from Andy Budd’s model that they had in Brighton. It’s really successful and we’ve get a lot of people coming along all the time.

I used to work in the university. I worked as a Web developer in the university, and nobody would come. And we were even sponsored by a university organization, and still, the lecturers would not come.

And I spoke to the organization that sponsors us about why this is, and their response was, and their specific remit, is to get people from education and digital media people working together. They specifically said, “They’re too arrogant. They think that your lot know nothing. You’re just the people out there on the ground, doing whatever you do, and they think that they’re better than you.”

[inaudible]

Jeremy: This is the problem with academia in general, I think. Okay. Yes, that is a problem.

Man: Speaking as a university lecturer…

Jeremy: You will wait for the microphone, Mr. University Lecturer.

[laughter]

Jeremy: I don’t care how many letters you have after your name, you wait till the oompa loompa gives you a microphone.

[laughter]

Man: Hello? Yeah, speaking as somebody with letters at the BEGINNING of my name… [laughs]

[audience groans]

Jeremy: Okay.

Man: As a university lecturer teaching Web standards, at bachelor’s level, at the University of Salford, I said this last year. And on the one hand, I have to agree: I am probably the only academic here. And that’s a shame, because there should be more of us. But there are others. I’m not the only one in the country.

Jeremy: Yeah.

Man: There are bachelor’s level university education benchmarks out there teaching Web standards. And if Patrick does want to bring us all together, it would be great.

Jeremy: I blame Patrick for everything.

[laughter]

Jeremy: Dan, you know it’s only a matter of time. Little Jack grows up, he goes to school, he comes home one day, he says, “Daddy, Daddy! I just learned how to use FrontPage.”

[laughter]

Jeremy: Alex knows what I’m talking about.

Dan: Here’s the scary thing, though: actually, I learned a lot using FrontPage. [laughs]

[laughter]

Jeremy: Okay.

Dan: I swear to God!

Jeremy: You know what I mean.

Dan: Yeah, that’s frightening.

Jeremy: That just kind of gets to the second half of the question: should we expect everyone to be self-taught?

Dan: I don’t think so. The fear is that there’s so many different ways to do things on the Web, right? There’s no real one way. And this is probably wrong, but whenever I think of taking a class in Web design, I fear that there’s sort of this one way of doing it that’s passed down. Jeez, did that answer the question? No, I didn’t answer the question.

But being self-taught myself, I think it works, right? But it doesn’t work for everybody. So, no. The answer is no.

Jeremy: Okay. And you’ve done your bit to pass on what you know, right? Blogging and writing books.

Dan: I’ve tried to, yeah.

Jeremy: I think people all recognize the need to try and do that.

Joe: Are we doing user-contributed questions?

Jeremy: We are. Should I start doing that now? We’ve got a couple of more that have been submitted beforehand, and then I’ll open it up. We’ve got quite a bit of time left.

Joe: Very good.

Jeremy: Let me run through the last few that have been written down.

Joe: Maybe we can do a lightning round.

Jeremy: Yeah, let’s do lightning rounds as well.

Okay. Oh, this is interesting. We’re talking about conventions, and Colly had many derogatory slides of Jakob Nielsen in his presentation. The question is: “At what point do the conventions of design trends become crutches for designers without an original idea?”

Ooh, I’m thinking gradients and drop shadows and rounded corners, at this stage. Yeah, Dan, I’m actually going to turn to you about this one.

Dan: Wow. “Become crutches.” That’s tough. I really enjoyed Jon Hicks’ presentation earlier, about looking outside the Web for inspiration. He was talking about taking a t-shirt design and actually applying that idea to the Web. I don’t know if I understand the question. “At what point do conventions of design trends become..?”

Jeremy: We want a point, a specific point in time, measured in minutes.

[laughter]

Dan: Oh, oh. I completely… Yeah, it’s the wine.

Jeremy: Yeah.

Dan: “At what point?” Well, that’s a tough one, because everyone’s saying gradients and rounded corners, and they’re saying, “What’s next? What’s going to be next?” And I don’t know what that is. I think we’re still limited by CSS and HTML and all that. Multiple background images. That’ll help right?

Drew: It will help with the rounded corners and gradient colors.

Dan: Exactly.

[laughter]

Dan: We’ll have even more rounded corners. Everywhere. I think that to a certain extent the limitations of designing for the Web have affected the design.

Jeremy: Yeah. Which is fair enough. Every medium I guess has its own something. But, Joe, do you sometimes wish there was more of a traditional grounding in classic design? Very positively, there’s been great talk lately about typography and grids, but it’s generally Photoshop tutorials: here’s how you do a gradient.

Joe: True, but we are constrained by the boxy-ness of HTML which has inline elements that have no shape, merely width, that could break over line, and block elements which are rectangular. CSS is based on a box model - a nonsensical one, but it is a box model nonetheless. So everything is rectangles, right?

So if you can manage to make things that don’t look like rectangles and aren’t rectangles with a fake beard tacked on. Like three sides are a rectangle, but flames on the bottom or something. For your heavy metal site, I don’t know. If you can do that, and you’re someone like Veerle Pieters, who manages to achieve a sort of girly-ness on her design which is a difficult thing.

If we’re going to be stuck with crutches, can’t I have correctly implemented crutches? If you’re going to have a left hand nav bar, and you wanted custom type, can you not do GIFs? Can’t you do sIFR and a proper unordered list and have it valid HTML and not three tables for that?

If we’re going to be stuck with crutches, can we have well manufactured crutches? I don’t know.

Jeremy: Drew, what kind of crutches have they got over at Yahoo! Right now?

[laughter]

Drew: Don’t ask me about design.

Jeremy: Okay, fair enough.

Drew: I guess I think what comes after rounded corners, I think maybe square corners.

[laughter]

Joe: Or three rounded corners and one square corner.

Jeremy: So square corners are the next rounded corner.

Joe: By the way, I actually sort of like that Web 2.0 look. I think it’s quite vibrant and punchy and even crap fonts like Arial look really good in them. I actually don’t mind it at all, and if it’s become a style, fine. There are tons of styles.

Jeremy: It somehow feels un-British to me.

[laughter]

Jeremy: So American. I’d like some more classic Swiss design.

Okay, moving on.

Now this is something that kind of follows on from my talk earlier, mashing it up with Joe’s talk from yesterday. “If we dismiss Ajax as not accessible”, which one could read from my talk earlier, I don’t know if dismiss is the right word. But Ajax, accessibility; we have big problems. “To what extent is this ‘not our problem’?” Including scare quotes right there — “not our problem” which is the phrase that Joe was using inside his presentation.

Joe: Yes, yes, and misrepresented by at least three people in the audience.

In any event,

Jeremy: Actually, I don’t remember asking Joe to answer this question. But, Joe..

Joe: You looked at me, Mr. Chair.

Jeremy: Okay.

Joe: If I may have the floor.

Ajax — who is dismissing Ajax as not accessible? It’s more accessible to a lot of people. It’s really the instant feedback you get. That’s great for people who have a hard time reading your acres of text. You can just click on something and have it just change.

When people say Ajax is not accessible, they mean, “It doesn’t work in JAWS.” Well, okay, whose problem is that? Exactly.

I dispute the premise that Ajax has been dismissed as inaccessible.

Gmail for example; the default view of Gmail may be a dog’s breakfast for someone using a current screen reader, and if you listen to the new Google Code podcast, Episode 1, you will be enraged at the basics of Web standards that they only figured out after they did expensive user testing instead of hiring actually competent developers for Google. You won’t believe it. Maybe if they’d have had competent people working on staff instead of boy racers with Aspergarian tendencies, then maybe the default view of Gmail would have been accessible in the first place. Perhaps it’s all about competent development. Right?

Jeremy: I don’t know. There’s technological barriers. It doesn’t matter how competent you are, some things just…

Joe: It’s true. It might be extremely difficult — okay, it would be unfathomable — to make an image editing application that lives on the Web accessible to a broad range of users with disabilities, especially mobility-impaired. That would just be impossible. But that’s only because the browser’s an impoverished environment.

I know a graphic design graduate in Toronto who has severe physical disabilities, and is in a wheel chair, and has very little muscle control, and he runs Photoshop, and has graphic design skills based on almost no movement of his wrist. So at the desktop level, even someone with that level of disability can still be a graphic designer. But Web browsers are impoverished environments for complicated things like that, on which you want to overlay things like accessibility, alternative input devices, and so on.

Jeremy: I know at Yahoo!, Drew, there’s a technological approach to try and solve these problems, with the YUI library. I’m not going to dig at Yahoo!, because here’s some really good work being done. Not so much in the code itself, the JavaScript code (you know, another library: great) but the documentation of the design patterns, and simply putting that out there, and saying, “You really need to be thinking about this kind of thing; you do need to be thinking about fallbacks.” It’s kind of good to see that at least being thought of in Yahoo!, even if as Joe points out, they’re not thinking about it in Google. I assume at Yahoo!, you do think of it as being your problem.

Drew: Yes. Taking a sort of pragmatic approach, you have to separate the big picture of, “Are products like JAWS deficient?”—the answer is “yes”—with the day-to-day “Are our users using products like JAWS today to view these pages?”—Yes.

So when it comes to actually implementing things, you do have to take account, and you have to scale back what you might like to do to what’s actually going to work, as much as you want JAWS to get their act together. And we all do want people like JAWS to get their acts together. I don’t think it’s fair as a community. We hold Microsoft to account for deficiencies in Internet Explorer, and then ignore less widespread products.

Joe: But you address that problem by doing shit loads of user testing, right? I remember Nate Koechley’s podcast from one of the other con’s, where he explained how they build Yahoo! Photos, which is now going to be killed. With any luck, that quality level of code will be transposed onto Flickr. There was massive testing involved to make it actually work with assistive technology.

That’s an amazing Ajax application — you can drag a marquee rectangle to select several photos, and then drag the photos into a folder. These sorts of things. And all that actually works, right? I think, if I’m not mistaken, Yahoo!’s solution is “test test test,” right? It should be everyone’s solution, really. Do I have that right?

Drew: It is a good solution.

Joe: There.

[laughter]

Jeremy: User testing is generally a good solution to just about everything, or at least a way of finding the good solutions. Test early, test often. User testing.

The problem with accessibility, though, because the way this question gets phrased: Ajax - not accessible. It’s one or zero. Yes or no. Current on, current off. It’s not that simple. Your case study for the IceWeb conference showed that. If you asked the question, “Is this application accessible?” The answer is always going to be “It depends.” You’re never going to get a straight yes or no answer. Did you want to talk about the results?

Joe: Yeah, I barely remember it now. Just Google my name and “ice Web” and you’ll find it. All my users could do a simple Ajax test in BaseCamp. Add an item, delete an item, modify an item. If you can’t do those few things, you can’t use the program. All my users, with a whole range of equipment and skill sets and so on, including one super expert, they could all do it. In a couple of cases, inconveniently, but they could all do it. That surprised the hell out of me.

Jeremy: So it wasn’t so much a technological thing, it was a learned behavior. They were able to figure out how…

Joe: I guess they were just so good at patiently plodding their way through things, and they were able to figure it out, yeah.

Jeremy: But whether it’s successful or not, well, it depends.

Joe: Yeah, and if we did a real test, and we compared non-disabled users, and this particular group of screen-reader users, and if it took them twice as long, and they made twice as many mistakes, then that would be a failure. You’d have to get it down to a smaller multiplication factor, but it’s never going to be 1:1. Maybe it should be 1.2, and one extra mistake, but it shouldn’t be twice as many, and twice as many mistakes.

Jeremy: Dan, I think Ajax introduces some design challenges, as well, because the browser is kind of taken out of the equation. Things like indicating something’s just happened, do you have to design for that now? Do you design spinning little wheels, and progress bars?

Dan: I’ve designed a few spinning wheels in my time. Yeah, it does present challenges. I haven’t done a lot of it, but yeah, actually the spinning wheel is probably all I’ve had to deal with.

Jeremy: Yeah, you let the developers worry about the Ajax…

Dan: Ajax is sort of out of my hands at this point.

Jeremy: Fair enough, fair enough. Richard, do you have any involvement with the ARIA group, at all?

Richard: No I don’t.

Jeremy: Okay. Interesting. Well, that’s really helpful.

[laughter]

Jeremy: The ARIA group — Accessible Rich Internet Applications — they really want to move forward to brace and extend HTML to have especially “role” and “state” namespaced attributes, so you could say this has the role of being a slider widget. This has the state of being the error message, or whatever. It would be really useful if screen readers could tie into that, and say, “all right, I know I need to watch this part of the page, because that’s where error messages are going to appear.”

Joe: Although, according to my reading, it requires XHTML 1.1. It’s not a new module for XHTML 1.0. And of course, as we know, 1.1 requires an XML mime type, and requires draconian error handling. I don’t know why they picked that. Rich Schwerdtfeger, whose name I’m mispronouncing unfortunately, who’s with IBM, is the mastermind behind this. He’s a certifiable genius. All the ideas are great, but I wonder if that little definition will be enough to sink it - I don’t know.

Jeremy: IBM and Mozilla together are working on some pretty interesting stuff there.

At this point, I want to hear what you folks have to say. Oompa loompa’s at the ready. [laughter]

Put up your hands if you have a question. Wait if the microphone gets to you. Keep your hand up so that the people with the microphones can get to you. This gentleman here in row number four on the edge, he’s an easy target.

Joe: And Mr. Chair, should we remind them that it’s Hot Topics?

Jeremy: Yeah, it better be a good question. Oh, we have all these books up here. They’re just for decoration. No…

Joe: We’ll be pulping them later. We’ll have a bonfire.

Jeremy: We’ll give you a book if it’s a good question.

Joe: Joseph Quinn.

Joseph: Can you hear me?

Jeremy: Yep.

Joseph: By using terminology like “not our problem,” is there a danger that people who aren’t on board with accessibility will just take that to mean “Oh great, we can blame it all on JAWS, and screw disabled people.”

Joe: The competent listener to my presentation, or listen to the ultimate podcast, or reader of my notes, will note that the title was “WHEN accessibility is not your problem.” It was not “accessibility is not your problem” - accessibility is your problem, it’s everyone’s problem. Hold on, I’m answering your question.

[laughter]

So if you weren’t listening, and weren’t paying attention, or you were tired and you missed something, and you don’t want to re-listen to the podcast, or if you missed it on the podcast, or if you didn’t read the notes properly, that’s really not my problem. I put a ton of work into making this thing very specific, and saying things like “cognitive disability guidelines may be inapplicable to certain sites, but that doesn’t mean other guidelines are inapplicable.”

Even on your personal blog, including your crappy LiveJournal, we can expect good HTML, and alt text on images, and this sort of thing. We don’t have to obsess about abbreviation acronym, because the world of written language contains multitudes, it cannot be encompassed by two elements in HTML. We can still use them according to semantics, preferably only when there’s confusion. Font resizing is strictly a browser issue. Those are the things I was saying, and in a nutshell, that’s how quickly you can sum it up.

So, it’s okay if people get the wrong impression. People get the wrong impression as I walk down the street. That’s fine, right?

[laughter]

Joe: But if you haven’t tried to listen carefully or read the notes carefully, then what am I supposed to do about that?

Jeremy: I bet you are really glad you asked that question.

[laughter]

Joseph: I’m not suggesting for a second that’s what you were saying, but what I’m suggesting is that some people, they’ll read the headline and they’ll say, “Oh, great. Joe Clark said accessibility’s not our problem.”

Joe: No, Joe Clark gives you a list of limited, edge case exceptions, when accessibility is not your problem, as a developer.

Jeremy: This has happened before, where people have written maybe controversial headings that can then be taken out of context. Jeffrey Veen had a blog post, “Why I don’t care about accessibility,” which was very attention-grabbing. His point was that when he hires developers, he expects them to understand accessibility, and he doesn’t specifically ask them, “Do you know about accessibility?” It’s just a given.

And Andy Budd gave a talk, “Why I don’t care about Web standards.” Of course he cares about Web standards. It’s more that, “Okay, we’re at the stage now where it should be a given, and if there’s dinosaurs out there who aren’t using it, well…”

Joe: Yeah, we seem to recapitulate this problem over and over again. Hold on. “Considered harmful” titles considered harmful… You should never begin the title of an essay with “To hell with…”

[laughter]

Joe: Martin Kliehm, this is for you. Right? Yeah. In the Web world, we seem to blow it with titles. And that’s okay. Maybe we should…

Jeremy: This happens on the blogosphere a lot, right? People write deliberately controversial titles, to get traffic and get Dugg no doubt. Fools.

[laughter]

Jeremy: But you do get a book for that, of your choice.

Joe: Yes.

Jeremy: Of your choice. Now, somebody else, I’m sure, has a question. And oompa loompa will bring a microphone. There we go.

Man: And so, you can do some cool stuff with Ajax, but if you were just going to use it for the illusion of performance increase or speed increase, then why not just optimize instead and stick with the accessible site?

Jeremy: Okay. So the question is Ajax can do cool stuff, but if you’re only doing it for the illusion of speed—which is kind of what’s at the heart of Ajax, I guess, the fact that you’re updating a discrete part of the page—why not optimize and do it without Ajax, but with leaner, meaner documents to begin with?

Yeah, that’s probably better, because that’s going to be probably more accessible. But in the real world, people are making these fairly rich pages, that a full-page refresh just really does seem to drag. And it makes the user experience so much nicer to have a discrete part of the page updating. And it does make it feel faster, more responsive, better, nicer. It gets people coming back to your site, when it’s implemented well.

But you’re absolutely right; then the accessibility question rears its ugly head. As I alluded to in my talk, there are some hacks we could be looking at—the tab index equals minus one combined with focus is one. And as I said in my talk, and I’ll repeat again now, Gez Lemon is the man to talk to there for some very good research on these things.

We’re kind of hacking around accessibility in Ajax. We’re trying to do our best with what we’ve got, because we have to. Hopefully, these are all stop-gap measures. The alternative is we don’t do anything; we just don’t do Ajax at all. It would be kind of a shame, because there are times when it would be good to use Ajax.

I think the question you have to ask whenever you decide to implement Ajax somewhere is what’s the core functionality? Because sometimes you can implement Ajax without affecting the core functionality.

A classic example is Google Suggest. You’ve all seen Google Suggest, right? Or you’ve seen those type ahead sort of searches on any blog with this Ajax form. The core functionality is doing a search: you type something in, you hit submit, and you get search results. The Ajax functionality is that, as you’re typing, it already starts fetching possible words that you want to be typing, to finish the words for you.

But the core functionality is searching. So you can add Ajax in such a way that it adds usability, makes it seem faster, does these benefits, without it being this yes-no, 1-0, current on, current off question of, “It’s inaccessible, simply because it uses Ajax.” There’s no check list approach when it comes to accessibility in general, and certainly not when it comes to accessibility in Ajax. It’s a case by case basis.

But it’s a fine question, and you win a book of your choice for that question. Did anybody else want to tackle that, or am I completely monopolizing the Ajax questions? Lots of hands up here. And you choose. You choose. Pick one. Make them fight to the death.

[laughter]

Jeremy: Gavin and Steve, fight it out. That’s very gentlemanly of you. Steve.

Steve: I was just going to more make a point than ask a question.

Jeremy: Well, actually, pass the microphone to Gavin, because we want questions, not points.

Steve: Okay.

Jeremy: I’m serious.

Joe: No grandstanding.

Gavin: Hello. What’s the environmental impact of what we’re doing?

Jeremy: Environmental impact of what we’re doing.

Gavin: We have different technologies that optimize the amount of bandwidth that we use when we build our applications, when we build our pages. We have different technologies that use more or less CPU. We should think about this as we’re building applications. What does the panel think?

Jeremy: That’s a very good point. The weight of the Internet at its last measurement is two ounces; all those electrons add up. The energy required to power those two ounces is 200 million horsepower right now. How much of that is Google, I don’t know.

[laughter]

Jeremy: This is a big question. Richard, you were talking about “Weaving the Web” as giving this big broad term picture. Is this something you think about, the environmental impact of what we do?

Richard: I have no idea at all, I’m afraid.

Jeremy: Great.

[laughter]

Jeremy: Drew, over at Yahoo!, do you recycle all your jellybeans?

Drew: I guess a positive impact is potentially in the reduction of production of things on paper and the traditional methods that we’re able to cut back on by doing things electronically. In terms of the impact of doing that electronically, I have absolutely no idea.

Jeremy: Joe, I bet you’ve thought about this.

Joe: No.

Jeremy: Really?

Joe: Never. Well, not per se. I have a lot of devices on in my house all the time: multiple Macs in sleep mode, a router and a cable modem, unnecessarily plugged-in telephones, features that only work on power, a stack of VCRs, Beta and VHS, a DVD player… it goes on forever, right? So in fact, just in stand-by mode, which is most of the day, I’m just sucking back a lot of juice for no discernable benefit. I have thought of that.

It would be very difficult, even if you went to four decimal places, to quantify the energy usage of, I don’t know, Gmail versus reading a LiveJournal. It would be very difficult to quantify that, I think.

Jeremy: There are some fascinating though somewhat boring stats out there about if most Web pages switched to having a black background rather than white, you might have energy savings that would mean in total for the electric grid.

Joe: On LCDs, I don’t think it makes a difference, because once they change, they remain there.

Jeremy: True. Good point. Dan, do you drive to work?

Dan: I walk.

Jeremy: Excellent. Well you’re doing your bit, clearly.

Dan: Yeah, I’m trying.

Jeremy: Good. It’s a good question, Gavin. You get any book you like. Steve’s still putting his hand up. Do you have a question this time? okay, let Steve ask a question this time.

Steve: Do you guys think that the fact that we seem to think of applications as having to be rich belies a lack of thought into the problems we’re trying to solve? I mean, could we not rethink the way we’re approaching these applications so that they can be enhanced with Ajax kind of techniques but not require them?

Jeremy: This is a good question. Everybody’s “Oh, rich experience, rich, rich, rich.” I hate that term rich Internet application, because it implies that if something isn’t a rich Internet application that it’s a poor Internet application. Like old fashioned documents are a poor experience. If there’s a rich user experience, there must then be a poor user experience.

Joe: And that leads to “skip intro.”

Jeremy: Yeah, okay, we won’t go there. But some of the richest experiences I’ve had on the Web have been on very plain document based sites. I remember coming across fray.com many years ago, one of the sites that made me realize I wanted to work on the Web. It was a document, it was a story, and it had an emotional effect on me. That’s a rich experience. These terms “rich experience” gets thrown around, it means…”like the desktop”: why are we in such a hurry to copy the desktop? I don’t think the desktop is that rich an experience, it just lets me open my browser, basically.

Joe: Yeah, and our stuff is already worse, because even if you don’t like Microsoft Word much, it doesn’t have a blinking ad for a casino in it, right?

[laughter]

Jeremy: Drew, at Yahoo! you do a lot of emulation of the desktop. It’s like you can’t half-heartedly copy the desktop. If you allow people to drag and drop in a Web application, then they will also be able to expect to use keyboard shortcuts. They will expect to be able to click and then shift-click to multiple-select items and things like that. If you’re going to emulate the desktop, you’ve got to do it completely. That’s a lot of work. That’s a lot of JavaScript, right?

Drew: Maybe. But I think the Web is increasingly replacing the desktop, for the majority of people’s usage. If you look at what kids are doing, they don’t necessarily even know the difference between the Web and the desktop; it’s all just stuff on the computer. You take away their Internet connection and they think their computer’s broken.

[laughter]

Drew: So I think it’s not a case of emulating the desktop, I think it’s taking some of the ideas that are good from desktop software, implementing them on the Web, and the Web, it just begins to replace a lot of those common applications.

Jeremy: I still see a lot of emulation, even in design. It seems like so many sites are just copying the interface of the operating system: the bevels, the way things look on XP or OS X. It just seems like such a lack of imagination.

Dan: Yeah, that’s true. That’s what people are familiar with, though.

I sort of am a paranoid person, but having all this personal information online rather than on my desktop… Maybe I’m old school that way…

Jeremy: You’re a fuddy-duddy.

Dan: [laughs] Exactly. For instance, people use financial application on the Web, right? Keep all your finances on the Web. To me, that scares the hell out of me. That should be on the desktop.

So I think things are going there, towards the Web. Someone mentioned Photoshop will someday be on the Web, a Web application. But it just seems so far away. I don’t know.

Jeremy: Yeah. This term “rich experience,” when they really mean “desktop-like experience…” I’ll tell you what’s a rich experience is a gigantic network of interlinked documents, using the simple technology of the A element. Hypertext: that’s rich. Wouldn’t you think, Richard? I mean, that’s really powerful.

Richard: I have an application which I use to look up Unicode characters and find out information about that. So yes, I have seriously nerdy moments from time to time.

I come across a lot of people who are trying to find out information about Unicode characters, and the problem that they always have is that somebody writes to the Unicode list and says, “I’m on a Mac. How do I find this character on the Mac?” And then somebody else responds and says, “Well, I’m on Linux, so I can’t really help you.” It’s a big problem.

And then I can pipe up and I say, “Well, I have a little application on the Web. Anyone can use it.” And so there are areas, I think, where it’s particularly useful to be able to have that sort of application type thing going on on the Web.

Jeremy: That’s a very good point, the universal access, as long as you have a browser: interoperability. Quick straw poll, just how you guys feel about things like Apollo, Silverlight—halfway stations between the desktop and the Web, but pretty proprietary, kinda closed. It doesn’t have that openness that the Web has, does it?

Joe: Tried and failed, and will fail again.

Jeremy: Fail, says Joe.

Joe: Come on.

Dan: Fail.

Jeremy: Fail.

Richard: Pass.

Jeremy: You’re saying it’s a good thing?

Joe: Abstain, abstain.

Richard: I’m abstaining.

Jeremy: Oh, that’s “pass” in the Mastermind sense.

Richard: Yes.

Jeremy: Okay.

[laughter]

Drew: Has fail.

Jeremy: Has fail?

Joe: Has fail, exactly.

Jeremy: Okay, fail. Adobe, Microsoft, you’re doing it wrong.

Joe: And those people who want rich applications, have they perhaps been looking at too many impoverished, static HTML pages that have poor graphic design? If they look at something by Bowman or Cederholm, would they think that is an impoverished Internet experience? I dispute that.

Jeremy: Yeah, this term, “rich,” to refer to desktop-like, I have a problem with it. There’s beautiful, rich experiences on the Web involving simple text documents. I say “simple,” but beautiful, emotionally engaging, wonderfully designed documents, linked together to form a World Wide Web.

Joe Clark’s announcement of his retirement from Web accessibility has been excised from the audio recording of the Hot Topics Panel

Jeremy: I would also like to thank my other panelists, Richard Ishida, Dan Cederholm, and Drew McLellan, Joe Clark. Thank them all very much indeed.

[applause]