On Tue, Jul 7, 2009 at 4:46 PM, Brion Vibber<br...@wikimedia.org> wrote: > Technically HTML 4 is pretty much the same in this regard; it's 100% > legitimate SGML and HTML 4 to skip implied opening and closing elements, > drop quotes on attribute values that are unambiguous, etc.
Not entirely. HTML 4 doesn't allow you to omit quotes on attribute values that contain non-name characters, for instance, at least according to the W3C validator -- so you need quotes for all URLs, for example. These aren't necessary either in practice, or in HTML 5. I'm pretty sure the requirements for opening and closing elements are stricter in HTML 4 as well. HTML 5 tends to loosen things up to whatever all browsers support, which is usually more lenient than what HTML 4 formally allows. It also actually specifies what constructs must be supported, in painstaking detail, so you can figure out what's legal without dumping it in a validator and hoping the validator's correct . . . On Tue, Jul 7, 2009 at 4:49 PM, Brion Vibber<br...@wikimedia.org> wrote: > At a minimum, I'm glad to see the dead-ended XHTML 2 working group > officially killed; actual compatible implementations of ongoing work are > happening in the HTML 5 world and that's where the future definitely is. > > > I don't see much need for us to stick with the XML formulation for the > next generation, given that we've never actually served our XHTML 1 > *marked* as application/html+xml for compatibility reasons: > > * IE refuses to display any content usefully > * Safari gets confused about character references > * even Mozilla will have different JS behavior, which would require us > to jump through some more hoops to kill the last document.write() calls... > * not to mention that your entire web site becomes inaccessible > instantly if you end up with a markup error in the page footer! > > Unless we're embedding our XHTML into other XML streams (which we're > not), there's little benefit to strictly sticking to the XML formulation > for page output. > > XML formulation could perhaps be useful if we migrate page text storage > from custom markup to an HTML-based internal format, as we could then > toss it at XML parsers without worrying. But that doesn't have any > bearing on the HTML user interface we display to end-users in browsers. Does that mean "go ahead and begin switching to HTML 5 now", or what? _______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l
