Why Don’t the Big 4 Switch to Web Standards?

It’s all about speed…. these guys optimize thier sites to reduce redundant code and increase content/byte.

See http://www.websiteoptimization.com/ for a discussion.

Outdated content-management systems that are too expensive to upgrade.

And no, I don’t think they all know about Web standards.

8. There are cases where implementing to web standards causes undesired behavior in certain browsers (looking at you, IE). Recently I worked on a frames-based site, and the *only way* to reliably suppress scrollbars in certain places was to induce quirks mode. It sucked, I felt dirty, but it was the only way to go.

I see no reason why it could not be done, from a technical knowhow point of view. These websites have been updating their content in various ways for a very long time and I think they like how it works.

They could definitely save some money by using standards based code, and I think it would pay for any development costs they would incur…

Big guys cater to millions of people world wide. Some of those people are still on IE2 or 3. According to The Browser May 2005 Stats there are still enough users using outdated software. And remember, if your site only servers 1000 users a month, then you might never get that user on IE2 or IE3. But big sites like Yahoo have way bigger chance of hitting that one user.

You can bet it’s all about the bottom line. You think someone like Google can hire possibly the brightest people in the industry and not have considered this? If you look at their source, every single byte in there is required for that page to work, nothing more, nothing less. When you factor in the kind of traffic they deal with, a few bytes here and there per page makes a HUGE difference.

I’ve seen a post from Yahoo looking for a web standards UI developer here in Sydney, Australia. So I think they are slowly moving towards that area. And obviously moving towards full web standards compliance will be hard to support old browsers and the increase in development cost outweighs the benefit.

Why should they? There is always two sides for any coin. Developers and information architects understand why it could make sense - but we live in our world, the business in theirs…

Googles business works now, why should they chance anything on their single point of access they have with their clients by changing to something that is a great idea but does not really yield benefits that translate into money.

To Joe, who thinks that legacy CMSs are the problem for Amazon, eBay, Yahoo, and Google: Your heart is definitely in the right place and support of open standards is clearly the right path to maintain as a direction and a goal of conscientious web application development. I am, however, surprised by your causal assertion given the well-discussed technical approaches for application development of each of the companies mentioned here… I assume at this point you know well that the content-management systems of the big 4 are not the problem. After attending a myriad of technical conferences (many of which you are also present for), isn’t it commonly known that each of the 4 mentioned can output markup, scripting, and CSS in any flavor currently supported by any user agent. And even flavors not currently supported, say, big chunks of CSS3, for example. Knowing these companies, your assertion strikes me as disappointing. Doesn’t it seem more accurate to assert that they’re choosing non-standard approaches explicitly?

To Nathan who thinks that they could definitely save money: I can understand your point, and it seems fair to say that cost reduction is easy to evaluate where scale is not a primary concern…however on what do you base your data for the unique developmental contexts and scales of these sites? Can you name, specifically, how any of these companies could save money by altering their presentation layer? It seems you’d have to know, in detail, how they construct their pages, their bandwidth and server loads, the number of servers, the browser/os breakdown of their user bases, the revenue generated by their current state of affairs, and what costs are incurred in their site construction.

The end point for me is that just pointing at big companies and crying ignorance or “legacy” doesn’t get us moved farther in a better direction. They’ve explicitly chosen (and continue to choose) a less-standard output and those of us who care should help raise the level of benefit that any company (including the biggest) would receive upon changing to fully-compliant output.

Or more succinctly…awareness and advocacy isn’t the only problem. As mentioned in this post - “I think that it tells us in no uncertain terms that for large e-commerce companies, the benefits of standards still do not outweigh the drawbacks.”

How do we help change this evaluation even after they’re aware of the many benefits of standards-based approaches?

Executive: Is it broke? Why fix it?

Web Standards Advocate: Well, sir/ma’m, these technologies, which we call “web standards,” are carefully designed to deliver the greatest benefits to the greatest number of web users while ensuring the long-term viability of any document published on the Web.

What the Executive hears: Blah sir/ma’ m Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah Blah.

Executive: ZZZZZZZ.

No, it doesn’t.

And I sign all my posts. Why don’t you?

There are web standards, and then there are web standards. I find it curious that the definition is that the site must have a doctype and a character encoding. For example, if you examine the source to the U.S. Yahoo! frontpage you will see a broken doctype and no character encoding. But then you will see a whole bunch of CSS, and not much in the way of tables.

Now that’s a much better attempt at ‘web standards’ than many sites you could examine. Personally I would be arguing that the lack of semantic content in the HTML is more worrying. But then, I wrote the European Yahoo! frontpage and tried to achieve more in semantic content (oh, and that page does have a valid doctype and an encoding), so I would say that.

The main problem is this: it is very easy to achieve ‘web standards’ on a page with only a handful of templates. It’s not on a site that has maybe 100-1000 templates. Then you multiply that by the number of sites that Yahoo! has. Then you multiply that by the number of countries that Yahoo! is active in. Then you divide that by the number of web developers Yahoo! employs, and discover a very small fractional number.

As for Google, this year, I (finally) set about fixing that wagon … in a manner of speaking. Read all about it.

Greetings! It’s almost a year later, and a lot more has happened on the Google front. Read Google Goes To Web Standardsville, Part Two for the scoop.