On 3/10/12 17:24, François REMY wrote: > | > Actually, they are both. > | > | Well, right, but there is no problem in my mind with making property > | _values_ case-sensitive as needed. > > If the property names are case insensitive, the way to reference them > should be case insensitive, too. Otherwhise, you will get weird > behaviors... > > { > a: lower-value; > A: UPPER-VALUE; > b: use(a); > C: use(A); > } > > I think the sanest solution here is to have identifiers be > ASCII-insensitives. Why only ASCII? Do we have reason to believe that making them Unicode case-insensitive (using the "simple" case folding algorithm, not language-dependent mappings) would be problematic? ISTM that, while less than perfect, it would be a big step forward from ASCII-only. JKReceived on Wednesday, 3 October 2012 16:50:59 GMT
This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:21:01 GMT