> What is the motivation for the F&O spec to define the > startingLoc argument as, xs:double and not an integer? Primarily backwards compatibility: XPath 1.0 only had one numeric data type, which was equivalent to xs:double in XPath 2.0. If it had been defined as xs:integer, then valid XPath 1.0 expressions such as substring($s, 1, (string-length($s) div 2)) would have been rejected as errors in XPath 2.0. Michael Kay http://www.saxonica.com/Received on Sunday, 23 November 2008 13:56:34 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 23 November 2008 13:56:35 GMT