Neil St.Laurent wrote: << That would make more sense, but then we definitely have a confusion as to which is more appropriate: /H1 P/ or /H1 //P// Would would appear to mean the same thing with your change to the selector semantics. >> I was looking for ways to get rid of the '//' altogether. Without it, contextual relationships can be described more logically. I would also like to see a change in semantics. Is it not the case that relationships between parent, child and siblings are all contextual relationships? So: ---- 1. Contextual relationships are relationships between an element and other elements in the document tree, and can be of two types: ancestral relationships (one element a descendent of another) and sibling relationships (two elements with a common parent). 2. An ancestral selector is made up of two or more selectors separated by white space and/or a tilde. 3. A sibling selector is made up of two selectors separated by white space and/or a tilde, and contained within a pair of forward slashes. 4. The tilde (~) narrows the scope of a contextual relationship. When the tilde is inserted between the relevant selectors, it restricts ancestral relationships to parent-child elements and sibling relationships to immediately-consecutive elements. ---- Which leaves the question of first and last elements. Suggestions include: 1. pseudo-classes; e.g. P:first, P:last, etc. 2. pseudo-attributes; e.g. P[first], P[last], P[2], etc. 3. pseudo-elements; e.g. DIV:first-child 4. another use of operators; e.g. DIV /~P/ (with the 'immediately-consecutive' operator alone implying that P is the firstmost sibling) <<I'm sure somebody would kill me if I suggested a more XSL like syntax for selectors at this point... :)>> Not me. I like to see possibilities. <<I'm thinking maybe we should let the guys who wrote the standard mull over this selector issue and get back to us on either a clarified or a modified proposal.>> Yep, this discussion has been almost as heavy as HSL nominations. David PerrellReceived on Friday, 5 December 1997 14:10:33 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:53:53 GMT