- From: Michael Kay <mhk@mhk.me.uk>
- Date: Sat, 28 May 2005 10:03:43 +0100
- To: "'Martin Duerst'" <duerst@it.aoyama.ac.jp>, "'Arnt Gulbrandsen'" <arnt@gulbrandsen.priv.no>, <public-ietf-collation@w3.org>
> XML Schema allows instances to distinguish between null values and > empty values. XQuery and XSLT are most probably using this. > Their WGs were in meetings the past week; I hope we'll hear > from some experts in the comming week. In my view a collation defines the ordering of a set of strings. The handling of non-strings (e.g. "null values" or other non-strings such as numbers and dates) that might need to be sorted along with the strings isn't part of the collation definition, it should be left to the host language. Different host languages have different data models and type systems, so it's best for the collation not to stray into this territory. Michael Kay
Received on Saturday, 28 May 2005 09:03:55 UTC