For example, recently a lot of energy has been spent on developing formal XML standards for office documents, but despite all the hard work and media coverage, the effort has met with only limited success. Microsoft released plug-ins that provided OpenDocument format (ODF) support for Office, but the files they output are pathetically incompatible with other software that supports ODF. Call that an anticompetitive move on Microsoft's part if you want, but the open source community isn't doing much better. OpenOffice.org claims to support Microsoft's competing OOXML standard, but its track record is almost as bad as Microsoft's. So much for the universality of XML.
The result? Microsoft Office's legacy binary formats remain the de facto formats for office documents worldwide. Microsoft wins, and open, standards-based XML loses. But who is really to blame? Is it Microsoft's fault -- or is a universal, XML-based office document standard that's supported consistently across every application suite simply an unrealistic goal?
Similarly, when it was first conceived, XHTML sounded like a great idea, but it quickly became the Esperanto of the Web: While its goals and ideals were laudable, its strict standards were simply too burdensome for non-engineers who were used to communicating in a less formal way. So where does that leave HTML5? If it specifies too much, it will suffer the same slow adoption rate as XHTML; not enough, and browser vendors will inevitably implement its specifications in inconsistent ways.
Get used to it. Improved standards compliance has been a boon to Web developers everywhere, but if we're honest, standards will only get us so far. I don't know how long it will be before the final HTML5 spec arrives, but if you're hoping it will usher in a standards-compliant utopia, I suspect you'll have to wait a lot longer. Maybe forever.