On Apr 3, 2009, at 09:50, Henri Sivonen wrote: > * A close approximation of the outcome (if I guessed your intent > correctly) could be achieved with a spec change with fewer > implementation changes if it were stipulated in a spec that > selectors don't match case-insensitively against HTML nodes but > match against HTML nodes after the selector ident has been ASCII- > lowercased. Furthermore, lowercasing the local name on createElementNS still wouldn't be enough to hide the actual implementation strategy of the appearance of selector case-insensitivity unless the createAttributeNode[NS] hole were plugged as well. (The HTML5 spec currently plugs the hole, but I think it's weird in principle and implementation-wise annoying if assigning an attribute node to an HTML element mutates the supposedly immutable internal state of the attribute node, which is why I filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=6689 .) -- Henri Sivonen hsivonen@iki.fi http://hsivonen.iki.fi/Received on Friday, 3 April 2009 06:58:45 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 3 April 2009 06:58:49 GMT