Buzz Archive: August 2004

Standards and Web 3D

Via /.: Former VRML heavyweight Tony Parisi has posted a thought-provoking article on the XML-based, ISO-backed standard for real-time-3D-over-a-network, X3D.

While X3D isn't one of our 'core' web standards, Tony has been working on standards — and on making them commercially successful — long enough to have some good insight on the relationship between standards, 'killer apps' and commercial success:

Standards don't matter because it's the applications that matter. … But try to scale these applications up, and try to reuse your content, across an enterprise or over the Internet, without standards. Just try. And even if you don't want to use standards, your customers will eventually make you, because by now they have gotten tired of paying you too much money to rewrite the same content over and over and over again for each new application use, each new platform.

His article also has some interesting thoughts on the history of VRML, why it failed, and why every other attempt at real-time 3D over networks has suffered the same fate. The comments also include some interesting ruminations on current standards, like SVG, and proprietary 'standards,' like Flash.

The impetus for Tony's post is the 3D Industry Forum's competing U3D format. The 3D Industry Forum is lead by Intel, and counts Microsoft among their members. They appartently intend to submit U3D to the ECMA, and ultimately perhaps the ISO, for standardization.

Heading in the right direction

Nearly two years ago, a new design of the Microsoft.com home page was met with near universal contempt from the web development community. Jeffery Zeldman took them to task for their gratuitous font tags, and Mark Pilgrim heavily criticised their terrible accessibility.

Fast forward to today, and the site has had another redesign - only this time, something's different. While the new design misses out on full standards compliance by quite a bit, the underlying code shows a dramatic improvement in both markup quality and accessibility. The table count is down to just 8 from over 40, the font tags have been relegated to the dustbin of history and the size of the homepage HTML has been slashed to less than 11 KB. This is a significant step forward, and one that deserves recognition.

A comparison of the new design to the old shows just how much of an improvement can be had through standards based methods. Compare Mark Pilgrim's snapshot of the 2002 redesign in a text only browser with this snapshot of the new design in the same. Web standards techniques are not an all-or-nothing approach; even incremental adoption can lead to substantial improvements. Hopefully Microsoft's new homepage is a sign of greater things to come.

For more on the new design, see Doug Bowman and Eric Meyer.

Web Standards a Campaign Issue?

No, not really. But lefty überblogger Kos gets in a good rant about them anyway:

Until browser developers learn to embrace web standards and allow for a uniform browsing experience, people like me will write code which will break on someone's browser, somewhere.

Of course, Kos is still slogging along in <table> land, and has some validation troubles to boot. But then he's a political commentator, not a web developer. Fortunately, he's got some clueful visitors who are helping him out.

The fact remains that there's really no reason why end users in 2004 should be struggling with browser incompatibilities in basic formatting. At this late date, we should have browsers that render standards-compliant HTML (looking at you, Internet Explorer) in a predictable fashion, and development tools that generate standards-compliant HTML by default (don't look so surprised, FrontPage).

(Not So) Heavy Chevy

Chevy has received glowing praise for cutting weight and adding polish to their new C6 Corvette. Now they've taken the same approach with their redesigned web site: it's both slimmer and easier to use thanks to semantically-sound, standards-compliant markup.

Oh, there's the odd typo here and there, but dig into the code and you'll be treated to elegant construction techniques that would do the Corvette engineers proud. And appreciation of the new site won't stop with the pocket-protector set.

Legend has it former GM president Alfred P. Sloan once said, "the primary object of the corporation … was to make money, not just to make motorcars." One expects that sentiment extends to web sites as well: the largest car division of one of the largest companies in the world isn't going to build a site just to show off their mad markup skillz. It's the bottom-line focus that counts, and the new site delivers. It boosts site reach by increasing accessibility while simultaneously reducing bandwidth and maintenance costs — a bean-counter trifecta.

The marketeers must be tickled, too, by the clean new look, trick expanding menus and aforementioned wider reach.

The Engineering, accounting and marketing departments all on the same page? How often does that happen?

Well done, Chevrolet.


Tip o' the hat to Deepak Mitra for the heads-up.


Update: A visitor by the name of Jack points out that Chevy sends users of Netscape 4 and earlier, Internet Explorer 4 and earlier and Opera 6 and earlier to a 'please upgrade' page.

Ouch. So close, yet so far.

As Jack says, using a JavaScript sniffer to 'protect' users of older software from an otherwise compliant site rather misses the point. At a minimum, the site should be served to those users as unstyled markup. Perhaps with a single upgrade warning, if absolutely necessary. While those users wouldn't get the sparkles and doo-dahs, thanks to the site's structural markup they would get a usable site with all the critical information.

Here's hoping Chevy patches that little niggle soon.


Update the 2nd: WaSP script wonk extraordinaire Porter Glendinning has pointed out to me that the Chevy site actually uses an object sniffer, rather than a browser sniffer per se.

Checking browser capabilities, rather than brand and/or version, is definitely a better way to go, though simply locking out browsers that fail the test is still sub-optimal.


Update the 3rd: Chris Moritz, the information architect on the crackerjack team that built the Chevy site has given me a bit of insight into why the site behaves as it does.

Turns out, the reason for the upgrade message is branding: Chevy felt that it would do more harm than good to let people on the site if what they got wasn't as polished visually as it could be. I disagree, of course. So does Chris, truth be told. But it's the client's brand, and the client's dime, and truth be told the number of visitors affected is small. So Chris and his teammates came up with a compromise solution — the same one Jack did, as a matter of fact: direct users of older browsers to the upgrade page on their first visit, but give them the option of continuing on to an unstyled version and set a cookie so they won't get the upgrade message on subsequent visits.

Speaking for myself, that seems like a reasonable balance.

Unfortunately, deadline pressures caused the cookie feature — and consequently the pass-through to the unstyled version — to be delayed to the near future. Chris assures me that it will be launched soon, along with other little improvements.

Here's betting it will be worth the wait.

Browsehappy: Bad Grammar for a Good Cause

The Web Standards Project launched on Friday a new campaign dubbed Browse Happy.

Despite Microsoft’s efforts to keep a competitive browser on the market, problems with Internet Explorer for Windows continue to mount. Meanwhile, Microsoft has announced that broad changes to Microsoft Internet Explorer for Windows will wait for the official release of the next-generation Microsoft operating system — scheduled for a date that is years away.

However, superior and secure alternatives to Internet Explorer for Windows are already out there. They all download just as quickly as the typical upgrade to Internet Explorer — if not more quickly — and will leave the Windows operating system practically untouched. The Web Standards Project has launched Browse Happy to make those alternatives easier for users, with the goal of making the web safer and more enjoyable for all who use it.

[A nod in closing goes to Jeremy Keith for his contribution to this post’s title.]

Accessibility Settlements by Priceline and Ramada

The New York State Attorney General’s Office published a press release this Thursday announcing that Priceline and Ramada had agreed to make their sites more accessible to visually impaired visitors:

Under the terms of the agreements, the companies will implement a range of accessibility standards authored by the Web Accessibility Initiative (“WAI”) of the World Wide Web Consortium (“W3C”), an organization that recommends Internet standards. For instance, graphics and images must have comprehensible labels, tables must have appropriately placed row and column headers, and edit fields (boxes where the Internet user inputs information) which must be labeled to indicate which information is requested. The companies must also implement a wide variety of other initiatives, based on guidelines authored by the W3C.

These settlements follow investigations over the last couple of years into the sites’ compliance with the provisions of the ADA, which the New York AG alleges that they must adhere to as “places of public accommodation.” It is important to note that, under the terms of the settlements, neither company admits that the allegation is correct, and they are making the changes voluntarily in order to prevent the state from filing suit.

Who knows? Maybe Priceline visitors will even be able to choose between William Shatner and Leonard Nimoy voices.

[And belated thanks to Chip Adams for the pointer.]

FrontPage Horrors

Frontpage 2003 is still the current version in August of 2004? I guess that’s the problem with using a date in your product name. Or maybe it’s an indication of a larger updating problem.

Here’s an article featured on the FrontPage homepage within the Office website: “Create a structured page layout by using layout tables and cells.” Wow that’s disturbing.

Read wg for more horrors… Can’t bear to read it and quote it anymore. It’s depressing.

SEO + Standards = ?

Well, that didn't take long. In his follow-up to his 'Silly Expert Opinions' post, Eric Meyer ponts to a test run by Peter Janes. The early results aren't necessarily encouraging.

My next question: in the comments to the compooter.org post, Shari Thurow insists that CSS does matter to Google. As proof, she points to a page on Google's site which reads, in pertinent part:

Hypertext-Matching Analysis: Google's search engine also analyzes page content. However, instead of simply scanning for page-based text (which can be manipulated by site publishers through meta-tags), Google's technology analyzes the full content of a page and factors in fonts, subdivisions and the precise location of each word. Google also analyzes the content of neighboring web pages to ensure the results returned are the most relevant to a user's query.

So how does that jibe with Peter Janes' findings? I've no idea. Perhaps a follow-up is in order: try testing text with CSS styling to make it bolder, bigger or different colors — including matching the page background as keyword spammers might do. Then also try using old-skool <font> tags to accomplish the same thing and see if either matters. Maybe try <i>, <b>, <em> and <strong> for good measure. Heck, why not try positioning and images with alt attributes while we're at it. I'd also be interested in results from Yahoo! and MSN Search, since neither of those uses Google's engine and both command healthy audiences, if nowhere near the size of Google's.

As Eric notes, it would be terribly disappointing (though not terribly surprising, I suppose) if SEO and standards were found to be substantially at odds. The only way we're going to know that, though, is to keep experimenting. Certainly Peter seems willing, as does Eric. And WaSP's own Molly Holzschlag seems interested in how SEO fits into the total client-education picture, if nothing else (note: Molly's post was written before Eric's follow-up, so she refers to 'slimy SEOs'; I expect she means that to be distinct from the honest, professional SEO experts who have been contributing to the compooter.org discussion).

Let's keep digging.

SEO Redux

Reading through the comments to the compooter.org article, it seems they contain the beginnings of an interesting, if occasionally heated, dialogue between the SEO and standards worlds. That's a pure good, IMNSHO.

If the message of the SES crowd in the comments is indeed what they were preaching at the conference it seems there's more than a little overlap between their concept of best practices in web design and ours.

Here's an idea: perhaps we standards folks and the SEO crowd should do a bit of knowledge sharing? In the comments, Danny Sullivan said he's already asked Eric Meyer to do just that, with an eye towards a possible speaking slot at an upcoming SES no less. That's a great start. But I think we can do more. I think there's gold to be found at the intersection of SEO and standards, or at least some good web development.

Let's keep the begining of dialogue in the comments to the compooter.org post, throw out the flames and ignorance, and use it to build a better set of best practices for web development. One that accounts for standards, accessibility, usability and search engines.

Maybe one of the new samaritans can redesign one of the sites linked above, or that of session moderator Chris Sherman's company, or panelist Dan Stone's company or the companion site for Shari Thruow's book (though that last is pretty close to the money as-is).

Or maybe we should work on some other site, one upon which nobody's livelihood depends. Use it as a test bed for accessible, standards- and search engine-friendly web development techniques.

Or perhaps just knowledge-sharing, as Eric and Danny are doing, will be enough to find the intersect.

Whatever. Let's just not let this opportunity go to waste.

SEO tomfoolery

compooter.org and Eric Meyer share a laugh at the expense of the panelists and organizers of SES 2004. Some samples of the SEO silliness, as paraphrased by compooter.org:

quotation removed due to doubts about accuracy

I must say, I've worked with a couple of SEO outfits recently and neither of them spewed any of that nonsense. Matter of fact, at least one of them recommended increasing use of CSS, more sophisticated use of semantic markup and so on. Oh, there were a few bits of advice that weren't terribly good from an accessibility or validation point of view, but on the whole they guys were knowledgeable and sensitive to the realities of assistive technologies and alternative browsing devices.

It's a shame SES 2004 was saddled with these clowns, rather than the more knowledgeable professionals I know are out there.


Update: Eric Meyer has pointed out via email that compooter.org might have been a wee bit harsh — if not seriously inaccurate — in his assessment of the SES 2004 session Advanced Design Issues: CSS, Javascript & Frames .

Eric cites the 'clowns' professionalism and generally quite reasonable postings in the comments to the compooter.org article. I encourage everyone to read them, paying particular attention to those by conference organizer Danny Sullivan and presenters Matthew Bailey and Shari Thurow.

Do feel free, however, to skip over the debate on whether SEO is just a bunch of snake oil and spamming techniques. As I noted above, it isn't — or needn't be, whatever the bad apples in the profession may think.

Most especially, it's worth noting Danny Sullivan's admonition that the point of the session wasn't necessarily teach optimal web building techniques but rather to teach people how techniques in use affect search engines.

Fair enough.

But what frustrates me a bit — and this, like the rest of the post, is my personal opinion only — is that the group seem needlessly tone deaf to standards and, to a lesser degree, acdessibility and other aspects of best practices. For example, Matthew Bailey defends his advice on browser sniffing by simply saying that if you use it, you need to update it as new browsers are released. Well, duh. But wouldn't it be better just not to use it at all? Maybe use progressive enhancement to support all browsers with the same pages instead? If you're giving advice, why not advise the best practice rather than just less-bad?


Update II: Eric Meyer has posted a follow-up to his snarky SES 2004 post. He offers a heartfelt apology to all involved for taking the SES crew to task based only on compooter.org's description of the session. In fact, it sounds like he's now really keen to speak at an upcoming SES. That's Eric all over: even when he arguably puts his foot wrong, he handles it in a classy manner and something good comes of it.

Netscape 7.2 Released

Via /.: The long-rumored update to the Netscape browser based on Mozilla 1.7, Netscape 7.2, was released today.

Personally, I much prefer the lightweight, browser-only Firefox variant of Mozilla, or even Mozilla itself. Still and all, it's nice to see Netscape alive, if only for nostalgia reasons.

Web Accessibility Toolbar for IE/Win

Chris Pederick's outstanding Web Developer Toolbar has long been a must-have tool for web developers & designers using Firefox and other Mozilla-based browsers.

Now, webheads who're still using IE for Windows (yes, there really are some, and they deserve our pity ;-) have a comparable tool: the Web Accessibility Toolbar. Much like Chris Pedrick's Mozilla toolbar, the Web Accessibility Toolbar allows you to validate pages, resize the browser, disable images or CSS and much more with the click of a button.

IE still doesn't have anything like the coherent and accurate script error messages one gets from Mozilla's JavaScript Console, let alone the functionality of the Venkman JavaScript Debugger, but this looks like a great tool nevertheless.


(hat-tip: Vincent Flanders)

A Web Standards Checklist

Once again, Max Design provides a great resource for the standards world: A Web Standards Checklist. The list is meant to help folks understand the breadth of standards and provide a tool for developers.

The list examines six distinct areas of interest as follows:

  1. Quality of code. This section hones in on DOCTYPEs, character sets, validation, hacks, structure and performance.
  2. Degree of separation between content and presentation. Here, the focus is CSS for all presentation.
  3. Accessibility for users. Multiple questions regarding accessibility including alt attributes, font sizing, and descriptive links.
  4. Accessibility for devices. Here, the checklist helps developers hone in on how well devices such as browsers, resolution, print, and hand-helds manage your site.
  5. Basic usability. Hierarchies, navigation, and consistency are included in this section of the checklist.
  6. Site management. Check for 404's, friendly URLs, and favicons, oh my!

The list is a must-have addition to any developer's toolkit.

You Said the 'S' Word!

Vincent Flanders apparently thinks 'standards' is a dirty word. Vincent takes WaSP member Douglas Bowman to task for not including a link to the finished work in his post about a table-free redesign of Microsoft's home page. He has a point. That is, until he flies off into bizarro world by critizing Douglas for 'linking' the topic of table-free web design to web standards.

Vincent apparently thinks Douglas is trying to say that web design is 'about' standards, that standards an end in themselves. The end-all, be-all of web design.

Quite understandably, Vincent objects to that view. He throws out a quote of his (actually, it's a quote of him quoting Jared Spool) explaining that businesses don't give a rat's backside about 'usability', 'design' or other buzzwords. Businesses care about things like revenue, expenses and 'shareholder value' (a weasel-word itself if ever there was one).

There's just one problem: Douglas never said web design was 'about' standards. In fact, the word 'standards' appears exactly once in Douglas' article. What Douglas wrote was 'graph after 'graph detailing the bandwidth and storage savings and reduction in development time and complexity his table-free redesign would achieve if adopted by Microsoft and how those benefits translate to reduced costs and reaching more customers more effectively. In short, how tableless design can help accomplish exactly those things that Vincent agrees are important to business honchos.

And what else would Douglas be writing about? He's a WaSP, after all. We aren't here for religious reasons. We're here because we believe standards have practical value for developers, users and — gasp — businesses alike. That was the pitch when we were founded, that's the pitch now.

So what's your point again, Vincent? Using the 'S' word once transforms an article on the business benefits of table-free web design into a proclamation that web design is 'about' standards? I don't buy it.

Even if Douglas had pounded the standards drum a bit more, would that have been so bad? If companies were as standards-phobic as Vincent, would they routinely proclaim their ISO ceritifications in their marketing materials as many do? It just doesn't add up.

A cynic might suspect Vincent of slagging a leading standards advocate and A-list designer in hope of starting a Jakob Nielsen-style crudstorm — and generating Jakob Nielsen-style publicity as a result. But I'm not (quite) that cynical.

So I really want to know: why do some folks go apoplectic when the word 'standards' comes up, even when they agree with everything else being said? Just why is 'standards' still a dirty word?


(hat-tip: Robert Scoble)

IE for WinXP SP2 Explained

Over on IEBlog, group program manager Tony Chor is lifts the kimono on the changes made to Internet Explorer in Windows XP Service Pack 2.

While the update isn't available for other Windows versions, and it doesn't offer any improvements to IE's standards support, the security enhancements made to IE in this update are highly worthwhile for users. They also may cause problems for some site developers, so it's worth your while to brush up on what's new.

The web has been buzzing of late with rumors that an interim update to IE was in the works -- something to tide users over until Longhorn. Well, this is it folks. The rumors of niceties like a download manager and a pop-up blocker all come to fruition in SP2. I don't know if Microsoft will release another upate or not, but the smart money says this is the last hurrah before the next bet-the-company Windows release in 2005 or 2006.


UPDATE: This CNET aritcle includes confirmation from Microsoft that IE won't get a 'real' update until Longhorn:

"At this time, there are no plans to release a new, stand-alone version of IE," a Microsoft spokesman said. "The current plan is to make new IE features available with major Windows releases...Aligning IE updates more closely with Windows releases benefits customers by minimizing the number of updates to deploy and service."

Don't Sweat the Small Stuff.

WaSP co-founder and emeritus Jeffrey Zeldman writes:

“I’ve seen people debate whether ‘leading’ web designers are all using the h1 header element exactly the same way on their personal sites. The question isn’t meaningless but it feels small and slightly beside the point. Likewise, the same ancient arguments about XHTML keep slopping to the surface. Don’t we have bigger water animals to sauté?”

The entry goes into other subjects, many of them personal, but it’s really about sitebuilding, standards, life, and how the three intersect.

The reason I’m grateful to be so closely involved with the WaSP is because standards improve the web, and the web improves the lives of the people who can use it. At least, that’s how it’s supposed to work, right?

Isn’t standards support about keepin’ it real?

Moz gets XForms

Via /.: the Mozilla Foundation is teaming up with IBM and Novell to implement XForms on the Mozilla platform.

WaSP Interviews: Jim Ramsey

Another brand new car smellin' WaSP interview is live as Jim Ramsey of the San Francisco Examiner talks about redesigning the site into a Web-standards compliant site.

Jim is hooked on standards. He discusses how he took the plunge, how the code stays clean for a large and frequently updated site, and how designers could influence the non-technical decision makers to embrace standards. Read the interview.

XHTML and document.write()

Ian Hickson has written a nice little explanation of the problems with document.write() in XHTML documents that are served with an XHTML MIME type.

Do DOCTYPEs Matter?

A short time ago, Anne van Kesteren and our own Molly Holzschlag had a brief back-and-forth regarding DOCTYPEs.

Anne makes the point that whatever DOCTYPE you use, all browsers will treat your markup as 'tag soup' (aka HTML) unless you send it with the correct MIME Type. That is, it's the MIME Type, not the DOCTYPE, that matters when it domes to determining whether a browser will treat a page as XHTML or HTML.

For her part, Molly notes that while an XHTML DOCTYPE may not kick browsers into strict XML mode, it does have an effect — sometimes a dramatic one — on how browsers render pages.

Both posts make excellent points, and are well worth a moment if you missed them when they were new.

Updating IE

Internet News has an article discussing a possible upgrade to Internet Explorer. I must say I don't put much stock into their speculation that IE7 may come out prior to Longhorn, the next upgrade to Windows. And some of the features they mention — like pop-up blocking and a download manager — will already make it into Windows XP SP2. Nevertheless, the article is evidence that the tech press is on to IE's standards shortcomings:

However, IE does not support CSS fully, which gives headaches to Web developers forced to write code for the majority of users.

Amen to that.

Also worth a glance over your morning coffee is the related /. discussion.

Washington, DC Web Standards Meetup

Thomas, Kimberly, and I are trying to light a fire under the monthly Web Standards Meetup for Washington, DC. Pay no mind to what the Meetup site says—we’re still getting together tomorrow night at 8:00pm at the Capitol Hill Capitol City Brewing Company (in the old post office building next to Union Station). If you’re in the area come on by and hang out. If you want us to be looking out for you drop a note on the meetup’s bulletin board.

New co-operation agreement between W3C and OMA

A few days ago, the W3C and the Open Mobile Alliance (OMA) announced a Memorandum of Understanding that will allow them to collaborate on mobile Web specifications.

This formal working relationship enables the two organizations to collaboratively engage in exchange of technical information and contributions. The result will benefit developers, product and service providers and others, by providing standardized technology at their disposal to accelerate the development and deployment of new mobile applications and services.

Is this good news, or is this good news?

(Hat tip: Kyle Barrow)