[Reiterates part of point #2 from http://lists.w3.org/Archives/Member/w3c-xsl-query/2009Nov/0075.html (Members only)]

Re the definition of "equivalence of two atomic values"...

In a definition whose "parameters" appear to be merely two values, it's odd to have one of the rules suddenly talk about a GroupingSpec. And while it could maybe slide by if 'equivalence' were only used for 'group by', it's now used by Switch too, so talk of GroupingSpec makes no sense.

I tried to find a solution involving just changes of wording, but it was ugly. (The problem is that a GroupingSpec might or might not specify a collation, and a Switch never specifies a collation.) So I think we should turn equivalence into an actual pseudo-function, with a third parameter, C, which is an optional collation.