> Another question is what the negative impact, if any, is when > returning say > xs:byte when xs:integer is the static type(apart from Per > Bothner's confusion > when investigating these corner cases). The only negative impact I can see is that some type errors won't be detected, for example if the value is passed to a function that requires an xs:byte. But that's true for any expression that returns a subtype of the declared type, or for example if the result of 1+1 is delivered as an xs:short - which to my mind is clearly allowed. Michael Kay http://www.saxonica.com/Received on Friday, 24 February 2006 18:13:12 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 22 July 2006 00:10:21 GMT