Google Chrome | Main | Why Auto Updating Browsers Matter

October 15, 2008

Opera: only 4.13% of the web is standards compliant

Ars reports on an Opera study that only 4.13% of websites are standards compliant.

While it is no surprise that very few websites are completely standards compliant, their methodology seems flawed. Websites are doing more and more dynamic javascript stuff after page load which could affect such stats. Also, major js libs are linked in via cdns, so it is unclear how useful their XMLHttpRequest usage stat is.

But I can totally verify that chinese websites love flash. I still have nightmares from the AOL China gecko testing days. Flashing, scrolling, floating ads are scary.

Posted by doron at October 15, 2008 6:51 PM

Trackback Pings

TrackBack URL for this entry:


Where is the weighting that shows how widespread the pages surveyed are in terms of popularity and overall page views?

There are tons of websites (orders of magnitude more than the tiny amount surveyed), but not all of them are vaguely important or interesting. If the right sites validate, the overall validation statistic is basically irrelevant. Nobody cares if some obscure blog doesn't render properly in quirks mode.

Posted by: Ben Basson at October 16, 2008 2:07 PM

I was worried that the original version of the MAMA writeup, which had its share of cross-referencing of documents in other parts of the writeup, might suffer from an incremental publishing schedule. This is one such case where I think it suffers.

There are some caveats and more methodology notes listed in the sections on scripting that would help clarify things here.

I spent a fair amount of time troubling over the issue of dynamically created content and how to treat it. In some ways, my current solution is still crude and first-generation. One thing MAMA can't currently do is read into scripts that are written by scripts. I realize that this has a high potential for skewing the results - especially an area like XMLHttpRequest which would likely go hand-in-hand with scenarios like that.

I did try to detect scripts that were written by scripts, but barring actually running it, MAMA currently can only sit back and be an observer that the situation is happening and not do anything about it. The truth is, I didn't know just how much MAMA would run in to dynamic/on-the-fly writing of scripts. Some of MAMA's detections were meant for archaeology and uncover where to take new features next.

Scripts-writing-scripts turns out to be quite popular (who knew?)...How much XMLHttpRequest is used in those scenarios will have to be an exercise for a future MAMA crawl.

Posted by: Brian Wilson at October 18, 2008 12:08 AM

Post a comment

Remember Me?