Liveblogging a presentation by Jeffrey Zeldman at @media2005, London, 2005.06.09 (@media; atmedia2005; atmedia; @media2005)
Julian Treasure (no relation) opens. “A conference about standards, but you couldn’t describe it as a standard conference” given our “stellar” lineup of speakers.
Jeffrey toured the U.S. with the Insect Surfers, we are told.
J the Z: “The Greatest Story Never Told.”
“Bujumbura is the capital of –” “Burundi.” “Burundi. It has to be said now.”
Shows an E-mail from 1998 that led to the founding of the Web Standards Project.
This is only the second Web-standards conference in the world, and the first in the U.K., which is impressive because we were told that Web standards would never go anywhere.
His bed ’n’ breakfast has wireless Internet access only for Windows 98. (“ ‘Wiffy’ is Joe Clark’s pronounciation of ‘Wi-Fi,’ which I think is much better… Half of my presentations are stolen from Joe Clark.”) “At a pure machine compatibility level,” a lack of standards is absurd.
At Waterloo Station, Jeffrey and his wife and baby took a footbridge on a walking tour. “If you have a baby in a pram, it makes you kind of wheelchair-accessibility-conscious, because it’s the same kind of problems.” And the hidden lift was out of service, but the other lift worked. “Imagine if you’re in a wheelchair and you’re starting at the other side,” using the first lift and getting stuck at the other end. One of the nice unintended consequences is that designing with Web standards means you become aware of accessibility and you start making Web pages that work for all kinds of people and all kinds of devices. But when we started it, our motivation was nothing like that; it was just angry laziness.
“If you have normal vision, I quit smoking six months ago and gained weight. If you’re blind, I’m 6′4″ and cut. That’s like having a separate Web site” for blind people.
In 4.0 browsers, there were four ways to script (plus noscript). Inconsistent tag support (e.g., Netscape’s invention of embed). “The W3C didn’t like it, so it never became part of any standard.” “Weak and inconsistent CSS support”: IE3 supported font-family, Netscape 3 no CSS at all. Even inconsistencies in table layouts. One early use of CSS was to trick Netscape and IE to treat tables the same way. IE5 beta was to have a proprietary DHTML (“Ajax”).
The early developers’s strategy was simple: We’re mad as hell. People will unite around problems, though they’ll disagree about solutions. Next: “ ‘Recommendations’ are for pussies.” “The W3C was kind of genteel.” It’s as though they were amateurs who were really happy to have done something interesting; “getting people to implement it” was a different story. “Netscape, which was on top, was certainly not going to join any organization that told them how to build their browser.” Ditto IE. Naming is very important: If you call these recommendations “standards,” that might change the way people thought about them.
Also they decided to make news. “As you know, you can actually influence media coverage of certain things you are passionate about or interested in.” Then they invited Web leaders, like Tim Bray, “much smarter than anybody else in the group – much more important than anybody else in the group.” It made the group look serious: “We have Tim Bray!”
Almost immediately, people started saying “This will never work.” Microsoft and Netscape have no interest in doing this; I’m tired of lost battles. That was a realistic defeatist attitude. Asking people to join and having people respond with how many ways it wasn’t going to work was something rather sad. “Fifteen people divided by a common goal” is how he describes the early Web Standards Project, since once you get people together the ways they disagree become clear.
Some people would rather fail grandly over a nuance than succeed by blurring something or being general. They formed an executive committee, and people had to react within a certain time if they objected to proposals.
By calling Web standards thus, “we thought that we had a chance of actually getting them onboard.” What to name the group? League of Concerned Geeks? Citizens for Conformant Markup? Rage Against the Machine? No, Web Standards Project (“because of Alan Parsons Project, of course”), since they assumed they would not be around forever. Plus WSP was similar to WaSP, and they’re creatures who work together but can also sting.
(“I used to work in advertising before I started working in Web development. That’s why I… well, that’s why I’m so old.”)
Their first phase was to protest. Just like buying timeslots so that the same TV commercial appeared on several stations at the same time, they had Dan Shaefer, Glenn Davis (Project Cool [remember them?]), Jeff Veen, Steven Champeon, and Dori Smith all published on the same day about WaSP.
They circulated a petition at Web Design World, Seybold, and other conferences.
Finally, “sheer ridicule.” “We’d write sometimes very indignant, sometimes very satirical,” sometimes “warm and fuzzy” diatribes about companies joining standards bodies and walking away from the standards. If you wanted to have additional nifty stuff that only worked in your browser, great, but support CSS1 and HTML4 correctly first.
If WaSP couldn’t be browser makers’ best friend, they’d be their worst enemy – and vice versa. So the CSS Samurai (Todd Fahrner, John Allsopp, Ian Hicks) listed the top ten problems in Opera so they can be fixed (later also IE5/Win). There wasn’t a top-ten report for Netscape because it was so bad at standards compliance, “there was nowhere to begin.” The one thing Opera got right showed that the spec was wrong (relative font sizing). “This made it seem as though we were not completely anti-Microsoft.”
In 1998–99, people left, and of course there was infighting, then inertia, malaise. In IE 5.5, CSS1 wasn’t supported, but coloured scrollbars were, and WaSP went ballistic. But because Microsoft was on trial about antitrust practices, anything about Microsoft was deemed newsworthy. But journalists never quite figured out that WaSP was talking about something unrelated to antitrust. Microsoft said it took them four hours to implement coloured scrollbars for a multi-billion-dollar client; we’re 89% of the way through CSS and you’re making fun of us?
1999: The Browser Upgrade Campaign “totally divided the Web-standards community.” “There are plenty of reasons to be against it,” but the thinking behind it was instead of testing for a browser, you test for standards compliance or capability. They assumed the campaign would only be used on Weblogs, “art sites and independent sites.” Also, nobody was doing really advanced CSS because they did not trust that enough people were using compliant browsers. “So we said: Pretend they are.” Let’s see what CSS layouts are really like. If they did, they’d find out how hard it was and what the browser bugs were.
Also in 1999: A Netscape 5 was supposed to be released, but it was going to be “a Netscape 4 with more patches.” WaSP persuaded Netscape to sit on its browser until the compliant rendering engine was ready.
1999–2000: Unsung heroes of Web standards include Fahrner, Tantek Çelik and IE5 for Macintosh. Zeldman told Tantek it would be great for accessibility if people could resize text even if it’s set in pixels. (People “shouldn’t” use pixels, but they would anyway.) Farher asked if he could show that to Mozilla, and Tantek said “sure.” Opera already had page zoom, and shortly IE5/Mac and Mozilla would have text zoom. Surely IE5/Win would have text zoom? “It didn’t work out that way.” “It’s hard to remember now how important that was,” but Tantek was rewarded by being transferred to WebTV.
Phase III: Help people some more, as with the Dreamweaver Task Force. Usually, offering free consulting to a company tends to get turned down, but Macromedia said yes in part because Jeff Veen had them as a client and arranged their first meeting with them.
2001: “The year of the compliant designer.”
Designers weren’t all that excited about standards until Dave Shea started the CSS Zen Garden.
2003: “Best-selling books came out in 2003, so basically standards broke.”
WaSP still exists, but during this whole time there were “independent people” who were promoting standards, “now much more so.” Don’t half the people in this room have a Weblog promoting CSS? A twelve-year-old boy had a site about his favourite CSS that he sold for $20,000.
The landscape today: “Thanks to Ajax, HTML is hot again. It’s OK to have HTML.” Thanks to Flickr, tags are hot. “You don’t have to be sort of ashamed to make an accessible site or use CSS. If you use CSS, it doesn’t mean you can’t design. If you say that you use HTML or XHTML, it doesn’t mean that you live with your mom and collect toy trains. It means you can be a professional in this environment.”
-
Question from Mike: At times, I’ve violently disagreed with what WaSP was doing, but two years on I completely agree with most of what you’re doing. So you were two years ahead of me! What will be happening in two years’ time?
A: Just because me and my friends were right doesn’t mean we know anything now. But using standards is being incorporated as a best practice. It’s a point of differentiation and can help sell you, but I think it’s going to be part of what is just expected, in the way that information architects are now expected on Web teams. Web standards are just going to be incorporated as a cost of doing business. “Designy firms doing Flash” as a separate category of Web developers may not last very long (Cf. Todd Dominey).
I hope the next version of IE will not mess up the bug-fix hacks we’ve all been using. (Mentions you can write directly to Dave Hyatt at the most secretive company on earth, and he writes back to say it’ll be fixed in the next version.) They started by blasting browser makers because they thought they didn’t have a chance with them, but ended up being friends with them. Even Chris Wilson of the IE team is “a big standards guy,” and the fact he couldn’t put everything he wanted into the browser may be office politics or company politics. Also, there may be bizarre surprises: The idea of the lazy, angry developer was not so that you’d have to know 732 workarounds to get your job done. What looks like a bug may be more-accurate rendering (as with transparent text in CSS3 in Safari).
-
Question from Dave, University of Salford: Do you think Adobe buying Macromedia will be an improvement? Will it have any effect on Macromedia’s support for strict DTDs?
A. I don’t know. Macromedia always seemed like a more accessible company to me. For Web people, Adobe always seemed to be missing out, but they had a different market (corporations). I expect they’ll throw some products out, but beyond that I don’t know what it means. Macromedia employees are writing in to him about the next version of Dreamweaver. Molly Holzschlag pipes in and says Dreamweaver will continue to comply with standards, and they have a friend at Adobe who is championing standards. There’s also a WaSP Microsoft Task Force.
-
Question from John at Inbox: About five years ago, they introduced a themes capability that took five times as much trouble to make work in CSS than in tables. They’re still fighting bugs this many years later. Do you believe it’s hopeful that I’ll be able to get to my retirement and not still be fighting the same bugs from the time I started Web development?
A: It does seem slow. It seems like an S-curve, with no help at all, then suddenly a lot of help.
[Scribe stopped writing at 2005.06.09 09:50.]