Mark Piligrim uses the phrase above in Dive Into HTML5 (a great, free book) to describe why certain specification or features have succeeded or failed in HTML. Pilgrim researched discussions on W3C mailing lists from 1993 to find how the <img> element came into being. The discussion in and of itself is interesting, but there was no magical argument for using the <img> element for images over the other proposed elements like <icon> or <a>. The only reason we are using the <img> element today is because Marc Andreessen shipped code with the <img> element before the others could ship code.
The W3C failed to “ship code” which ultimately lead to the death of XHTML 2 (eXtensible HyperText Markup Language) and its use in HTML5. XHTML is an XML markup language that extends HTML. XHTML 1.0 was released in 1997 and it included Appendix C which allowed XHTML 1.0 code to be used on documents served with the text/html MIME type. This was very important because the MIME type for XHTML documents, application/xhtml+xml, used draconian error handling. The text/html MIME type did not display errors in HTML to the end user. It simply fixed them as best it could. XHTML draconian error handling was different in that if there was an error it would be displayed to the end user and the page would just stop processing. This was a problem, a big problem. It was estimated at the time that 99% of HTML documents had errors in them. This means that without Appendix C 99% percent of the web would have failed.
Later versions of XHTML did not include this loophole where web developers could include XHTML on their web pages but not serve it with the application/xhtml+xml MIME type. It essentially eliminated backwards compatibility which is bad for the loosely formed and distributed nature of the web. The WHAT Working Group eventually broke off of W3C in 2004 over this issue. W3C insisted on inlcuding the later versions of XHTML in the HTML specification with would have made 99% of the web unusable. The WHAT Working Group thought it would be better to include more advanced features for forms without using XHTML because of the backwards compatibly issue.
So what do you think won? Backwards compatibility or draconian error handling?
XHTML 2.0 remained in draft limbo for several years. The W3C never shipped their code! XHTML 2.0 was never implemented by any browsers and the working group was canceled in 2009.
I think this experience heavily influenced the WHAT Working Groups decision to drop the “5″ from HTML a few days ago. They understand that HTML should be a “living document” that is always shipped.
