30321 2018-12-01 14:38:22 +0000 [XSLT30] (editorial) Section in fn:document explaining atomization of nodes assumes a single node can be atomized to a sequence of more-than-one, this is confusing 2018-12-02 20:47:32 +0000 1 1 1 Unclassified XPath / XQuery / XSLT XSLT 3.0 Recommendation PC Windows NT RESOLVED WORKSFORME P2 minor --- 1 abel.online mike public-qt-comments oldest_to_newest 129389 0 abel.online 2018-12-01 14:38:22 +0000 The section in fn:document (20.1), the second bullet point, reads: "For an item in $uri-sequence that is a node, the node is atomized. The result must be a sequence whose items are all instances of xs:string, xs:anyURI, or xs:untypedAtomic. Each of these values ...." The text starts with "an item ... that is a node". If a node is atomized, it will typically lead to a sequence of one. If the typed value is a list type, it may return a sequence of more-than-one, but this requires a schema-aware processor. I find this confusing. While the text is not technically wrong, I propose that in an upcoming erratum we clarify this, for instance by mentioning the exceptional case that can lead to more-than-one. 129390 1 mike 2018-12-01 17:40:19 +0000 I think that the spec should define what happens in the general case, and should not concern itself with the fact that for many users, the general case doesn't arise very often. 129392 2 abel.online 2018-12-02 20:47:32 +0000 Yes, you're right. Let's close this one, the list of bugs is already long enough and this appears to be not really a bug ;).