On Wed, 30 Mar 2011, Lachlan Hunt wrote: > > The spec states for submission of datetime controls that the value must > be expressed in the UTC time-zone. It's not clear whether this requires > the formatted string to state the timezone as an uppercase "Z" or > "+00:00", or whether either is acceptable. > > Most of the examples illustrate the use of the "Z"
I couldn't find any examples of the output of this; did I miss some? > but elsewhere in the spec, following the definition of the valid global > date and time string, it recommends against using the Z. > > "The *best representation of the global date and time string* datetime > is the valid global date and time string representing datetime, with > the valid time string component being given in its shortest possible > form, with the last character of the string not being a U+005A LATIN > CAPITAL LETTER Z character (Z), even if the time zone is UTC, and > with a U+002B PLUS SIGN character (+) representing the sign of the > time-zone offset when the time zone is UTC." > > http://whatwg.org/C#best-representation-of-the-global-date-and-time-string > > (Note that the spec defines this "best representation..." term, but > never actually uses it anywhere. It's not clear what the purpose of that > definition is.) It's used in the microdata to Atom conversion algorithm. > Opera has always submitted the value with a "Z" and not +00:00, and we > would prefer not to change this for backwards compatibility and ease of > parsing on the server side. The spec should define in the value > sanitization algorithm for the date and time time state, to express the > UTC time-zone using a "Z". Done. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'