Weblogs: Web Standards

@Media 2005: Zeldman keynote

Thursday, June 09, 2005

I've previously heard two Zeldman keynote speeches online. Both compelling listening. Watching him in person is a far better experience. He has the natural ability to win over an audience. Quiet, but engaging and well spoken. He's earned his reputation as the father of web standards, he's obviously proud of what he's done, but there's no trace of ego. Yes, seeing him in person has raised my opinion of him immensely.

Zeldman's keynote covered the story of the Web Standards project - an organisation he was instrumental in creating. He talked about the thoughts, strategy and direction right from the early days, and how the web moved from a position of multiple incompatible web languages and technologies into the far more coherent and consistent web standards approach. His slides are available.

The story of Hungerford Bridge

Zeldman retells his story about his visit to Hungerford Bridge, and provides an apt metaphor to web standards. The entrance up to Hungerford Bridge consists of 250 steps upward. With a pram that is pretty difficult going, but he manages it, and walks across it with his family to the other side, where he's faced with another 250 steps down to ground level. This time he is fortunate to meet someone who tells him that there is an elevator - great for wheelchair use and prams. So they take the elevator down to ground level. Coming back, they take the same elevator up again, walk back over to the other side. Of course, if one side has an elevator, so does the other. So Zeldman pushes the elevator button and waits twenty minutes. At this point he realises that the elevator this side doesn't work - you just want to shoot yourself in the head. Now relate that to the feelings a wheelchair occupant would have in the same position. Its just silly.

Motivation for web standards

As the use for web standards increased, developers started to become aware of accessibility. Nowadays the motivation behind web standards is accessibility - and its benefits, but it didn't start out that way.

The primary driver for the creation of a web standards advocacy group was anger. As Zeldman relates, anger is a powerful motivating factor. Originally the Web Standards Project was about angry laziness - there's absolutely no logical reason a web designer should need to support 17 or more different and incompatible implementations of dynamic scripting.

Getting started in anger

In June 1998 the NoEnd mailing list was started. This was in the days of the version 4 of Internet Explorer and Netscape Navigator. There were basically four different ways to script in these browsers (well, five if you include supporting no scripting). There was inconsistent tag support (e.g. embed for Netscape which wasn't supported in Internet Explorer 3). There was weak and inconsistent CSS Support. Internet Explorer 3 first supported font-family and font-color. We got promises of proper CSS support but that remained incomplete. While we had to deal with inconsistent table layouts, and then came Internet Explorer 5 beta with proprietary DHTML.

The Internet Explorer 5 beta sparked of real anger in developers, which lit of a series of conversation threads over email, involving Martin Diekhoff, George Olson, Eric Wolfram and others. Zeldman points out that the one big motivating factor was the anger. "Mad as Hell" was the strategy. People unite around anger and problems, but don't tend to unite around solutions.

The strategic approach

The W3C was seen as genteel. They had a brilliant specification for layout, but this wasn't implemented or supported by its own members. As Zeldman bluntly points out "Recommendations are for pussies". The Web Standards Project took those recommendations and promoted them as standards.

The second prong in their strategy was making news. If people paid attention to news, then there would be a chance.

Another piece of the strategy was to invite web thinkers and leaders, like Tim Bray, the co-author of the XML specification. Even though these leaders are too busy to play an active role in the organisation, their presence and input has a positive impact.

The idea was proposed and discussed. Inevitably discussions centred around the "it will never work" because the active web thinkers were at that point tired of lost battles. "Microsoft has too much control and won't support it". WaSP's approach was to find like-minded people and get them to list why it won't work.

To manage the WaSP organisations, they set up an Executive committee. Because committees suck - they get bogged down in cyclical discussion, disinterest and a large number of people, the Executive committee would be made up of three people. They would propose something, and if there were no serious objections within the following 24 hours, they went ahead and did it.

Creating a standards platform

The first move was to name the technologies around web design and call then web standards (even though they technically were not standards". That way the initiative had a chance to succeed.

Naming the group raised some 'interesting' suggestions including "Citizens of Conformant Markup". They decided on WaSP because the web design environment was a mix of like minded people, like a colony of wasps, and they sting when angered.

Phase 1: Protest

In media advertising there is a concept called a road block. During the time when there were only three TV channels, advertisers would buy an advertising block at the same time on all the channels, so they could air the same commercial simultaneously on all the channels. WaSP did that by publishing simultaneously on Builder.com (Dan Shaefer), Project Cool (Davis), Wired (Jeffery Veen), A List Apart (Zeldman), Webdesign-L (Steve Champeon), and Wise Women (Dori Smith).

The simultaneous announcements were intended to push the message to the browser manufacturers. One can imagine Bill Gates seeing an article on Wired about the anger of web designers coalescing into the web standards movement and thinking that this would be terrible for his company. Then switch over to Builder.com and see the same story - the simultaneous announcements could be enough to get Bill Gates to spill his coffee.

A petition was run asking web designers questions like how they felt about supporting five different types of scripting.

Ridicule was also a successful campaign, especially against organisations that are part of a standards body, but don't support the standards they produce. It needs reminding that web designers do like extensions and new features, but fist support the basic standards.

Phase 2: Assist

The CSS Samurai took up "The Cable Guy" strategy by being the browser vendor's friend or their worst nightmare. In order to have credibility friendship was important, highlighting when vendors did support the web standards. Todd Fahrner, John Alsopp, Ian Hicks, Eric Meyer and other CSS geniuses became the 7 CSS Samurais. (One later left, so they became just the CSS Samurais). The top 10 problems in Opera were identified, they told Opera who then fixed them. The same with Internet Explorer 5.

Netscape's support of CSS was so bad there was nowhere to being compiling a top 10 list. The only thing they got right was the bad relative font-size which showed a problem with the CSS specification.

By targeting a wide range of modern browsers, it showed the CSS Samurai not to be just anti-Microsoft.

Rocky times

Davis left but later rejoined. Infighting broke out (committees suck), inertia took hold. The malaise was highlighted by the coloured scrollbars incident. WaSP didn't care about what was added to the browser, but cared about supporting standards first. Internet Explorer 5 didn't get CSS1 support right, but they introduced coloured scrollbars. The media published wide coverage of this anger - it came at a rocky time, Microsoft were on trial for anti-trust behaviour, so the coloured-scrollbar incident became part of the big pylon of the anti-trust coverage. It was undeserved press attention since they were not talking about the same topic (anti-trust versus CSS1 support).

WaSP asked for CSS 1 support and got coloured scrollbars. Microsoft said it took 4 hours to develop coloured scrollbars, and it was requested by a big client. WaSP wasn't talking about the CSS improvements, so why must we listen?"

Browser update campaign

The browser update campaign divided the web design community. The thinking behind it was about avoiding the need to sniff for every browser, but rather check for feature compliance. Nobody at that time was using CSS layouts and semantic markup because it was not supported in the browser.

No-one was writing or trying out advanced CSS, or using semantic markup. So we asked developers to pretend they were writing for a CSS supporting browser. They would find out how hard it would be, run into problems and they either try to work around the problem or report it as a bug. This made people feel it was safer to use these techniques.

When Netscape decided to ditch the Netscape 4 codebase and elect to start afresh with a new standards based and open source rendering engine, it was tempting to declare victory for web standards. For a long time it looked like Netscape would be releasing a new browser that didn't support the web standard, but a petition was launched to persuade Netscape not to release a browse without supporting standards.

1999 - 2000

Tantek Celik was the man behind Internet Explorer 5 on the Macintosh. The IE5/Mac beta was great for accessibility since it could resize text even if specified in pixels - this was the Text Zoom feature. Microsoft allowed Tantek to show the feature to Mozilla - their arch competitor, and they implemented it too. Opera already had the Page Zoom.

The year 2000 was the year of the compliant browser - Internet Explorer 5 on the Mac. As a 'reward' to Tantek, Microsoft took Tantek off Internet Explorer 5 and moved his to WebTV. That didn't work out. Internet Explorer 6 took the next step in being more compliant.

Phase 3: Assist some more

The Dreamweaver Task Force was set up, led by Rachel Andrew and Drew McLellan worked with Macromedia encouraging them on building tools that can be used to create standards compliant websites. Companies typically refuse free consultancy, so Macromedia paid about $50,000 on consultancy fees. Jeffery Veen was influential making this happen.

CSS Zen Garden by Dave Shea was instrumental in getting web designers to embrace web standards. Best selling books were written, and bought.


WaSP still has a role to play, but it is no longer the focus. More people are making sites about web standards, and the movement is growing. The landscape today is about AJAX and tags. HTML is cool again because of these ideas. People are excited about XML and tags. There's no shame now in CSS and accessibility.

Questions and Answers: Two year prediction

I asked the question where we would be in two years:

Web standards would be a best practice, like usability. It can and will be used as a competitive advantage, and as a point of difference. There already is an expectation that freelancers understand IA, and clients will expect the incorporation of web standards as a way of doing business.

Flash experts will be expected to be conversant with CSS. Gone are the days where Flash was positioned to take over from HTML.

We hope that the next version of Internet Explorer will work out the bugs without messing up the bug-fixing hacks we need to employ today. Hopefully more openness and transparency around Internet Explorer 7. We started with blasting companies because we knew we wouldn't get their attention otherwise. Now they are listening - Chris Wilson, part of the Internet Explorer development team is a big standards guy.

There probably will be bizarre surprises. The idea of lazy developers not needing 732 workarounds is a good step. Getting browsers to the last degree of compliance is important. Safari already supports the CSS3 transparent text, so what looks like a bug might not be a bug, but a more accurate rendering.

Other answers to questions

Regarding the Adobe purchase of Macromedia, this might be an improvement. Adobe may throw some of their redundant products out. Molly affirmed that Dreamweaver will continue its support of standards.

The progress towards web standards seems painfully slow, but its more like an S-curve. It started out with no help, then a big jump in support which then slacks off. Things are getting better, because they have to. More sites are using CSS. Some developers made lots of money being able to support 17 types of standards and scripting technologies - web standards have made building websites easier for mainstream web designers.

Zeldman reflects on certain commentators have accused WaSPs position on web standards as creating standards only they could follow, making it too difficult for normal web designers to create websites. It is clear the complete opposite has happened, and this reduction of incompatible complexity has been the WaSP goal.

On using AJAX to extend browsers, browsers are powerful enough to create desktop-like applications. However we need to be careful not to break with the assumptions of how a web page works when dynamically changing a document. We must be mindful of usability and accessibility, AJAX applications breaking the back button creates the same range of problems as frames. There are new areas of expertise, such as figuring out how to mark out where people are in a page of an AJAX application (state problem).

On dealing with clients: Clients are the audience for Zeldman's book. We need to be showing the benefits and tradeoffs of web standards versus the old-school methods. Perhaps just give content editors a simple Content Management system - a blogging tool like Movable Type. Write styleguides - dumbed down about what they can and can't do. Tell people the really simple stuff like using <hX> tags for headers, showing them the source and the rendering. Try to preserve design and markup, and the integrity of the code. Factor into the budget costs for hold-holding support, and point them to articles about the benefit of web standards.

On search engines and hidden content. Will search engine penalise standards-based techniques like image replacement. So far they haven't, and hopefully they would look at whether a site is being deceptive with its use of image replacement before banning a site. Using semantic markup is certainly not punishable, and we can't see search engines doing that.

On the Acid 2 test, Molly response by saying its going really well. David Hyatt - working on Safari - has fixed all his bugs. Chris Kaminski reports that Konqueror and iCab pass the test. Opera is making good progress. We don't expect Microsoft to pass this test, since there is so much work to be done.

On .net generating extraneous and invalid markup. Molly again reports that a Microsoft task force will be announced very shortly. This will not be limited to browsers but cover all their products. Already changes have been made to Visual Studio. It is coming, it will take more time. BrowseHappy.com was given away to facilitate working with Microsoft.

Other coverage

[ Weblog | Categories and feeds | 2011 | 2010 | 2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003 | 2002 ]