This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
The phrase "valid schema" appears several times in Schema structures, and in particular is used in normative constraints. The term is not formally defined, and thus there are no hyperlinked terminology references. In principle, I would suggest that the term be given a formal definition and linked. In practice, I observe that some (perhaps not all) of the references are from composition-related constraints that are known to be controversial in the group. For example, in "Schema Representation Constraint: Inclusion Constraints and Semantics" [1] we find: "It [I.e. the URI that is the value of a schemaLocation attribute on an <include> ... NRM] resolves to a <schema> element information item in a well-formed information set, which in turn corresponds to a valid schema." We know that there is disagreement in the group as to whether it is best to say that such a <schema> element will in general correspond to a schema at all, much less a valid one. This bug is being opened not to restart that debate, but merely to point out that if the phrase "valid schema" occurs, it should be defined. Whether we can fix that without first resolving the larger issue, I'm not sure. See also bug 5164, which mentions confusion about the term "valid schema". (Bug 516 was discussed at the January, 2008 Schema WG F2F meeting. During the course of that meeting, I was given an action to open this new bug.) Noah Noah [1] http://www.w3.org/TR/2004/PER-xmlschema-1-20040318/#compound-schema
A review of all the occurrences of the phrase "valid schema" suggests that what is meant is, in all cases but one, "conforming schema". I suggest we change the usages appropriately from "valid schema" to "conforming schema". For the record, the occurrences are in: 3.8.4.1.4 Multiple Paths in Groups 4.2.2 Assembling a schema for a single target namespace from multiple schema definition documents (<include>) 4.2.3 Including modified component definitions (<redefine>) 4.2.4.2 Providing Hints for Schema Document Locations The exception is the occurrence of "valid schema" in the comment at the beginning of the schema for schema documents in K DTD for Schemas (non-normative) Here, "valid schema document" appears to be meant; I suggest we change "valid schema" to "valid schema document". From both lists above I have omitted occurrences of the phrase "valid schema document", which seems to me to be (a) what was intended and (b) well defined, modulo the problem reported in bug 5512.
Michael Sperberg-McQueen writes: > The exception is the occurrence of "valid schema" > in the comment at the beginning of the schema for > schema documents in > K DTD for Schemas (non-normative) > Here, "valid schema document" appears to be meant; > I suggest we change "valid schema" to "valid > schema document". Just curious: why "valid schema document" as opposed to "conforming schema document". As best I can tell, the term "Valid Schema Document" occurs in only one other place in the Rec, which is in a note and thus non-normative. To my eye/ear, "valid" has a hint of being valid per the S4S (to the extent one can actually validate against that), but seems a bit ambiguous on meeting other constraints. No big deal, but I was curious about it. Thank you. Noah
(Answering comment #2) The sentence in question, which appears in the opening comment of the DTD for schema documents, currently reads: With the exception of cases with multiple namespace prefixes for the XSD namespace, any XML document which is not valid per this DTD given redefinitions in its internal subset of the 'p' and 's' parameter entities below appropriate to its namespace declaration of the XSD namespace is almost certainly not a valid schema. The point being made here is that any proper parameterization of the DTD will "almost certainly" recognize a subset of the documents valid against the schema for schema documents. The additional XML representation constraints are not relevant here, and the use of the phrase "conforming schema document" instead of "valid schema document" would obscure the point. The comment echoes a remark in the introductory paragraph of the section, which uses the phrase "valid schema document". I hope this clarifies why the term "valid schema document" is proposed instead of "conforming schema document".
> I hope this clarifies why the term > "valid schema document" is proposed > instead of "conforming schema document". Yes, my mistake. Thank you. Noah
The proposal made in comment 1 was adopted on today's WG call, resolving this issue. Noah, since you were present your assent is documented, but if you would do us the favor of closing the issue, for the record, it would be good.