| |
Comment: |
Folks: The following from Vivian. I don't think these should hold up PC. She makes comments on both documents, but they are linked, so I put them both
here.
Dave,
Just spent some time reading the specs. It might be a bit late, but since I have read it, so might as well drop a few lines about the RNS-spec-v1.1 (
btw, the file name is still v10).
In general there are a few places that I felt are inconsistent:
section 1.1 page 3, “but a pseudo-schema for the port types “ [comment]: there is no pseudo-schema at all through out the document, only pseudo-UML,
so either pseudo-schema should be provided, or just say pseudo-UML, which is kind of strange in an OGF/OASIS document.
the name “RNSResponseEntry” sounds rather strange, isn’t it “RNSEntryResponse” is a natural word in SOAP?
All the “Example SOAP Encoding of the xxx Message Exchange” are up to <s11:Header></wsa:To> only, not sure it is the problem of using Mac Word
reading it, or it is authors’ intention – it doesn’t make much sense of having an example at all if it only shows a header.
section 3, “Requirement Level”, not sure “MUST” and “MAY” is a formal way of defining it, “Mandatory” and “Optional” are more commonly used
in this occasion.
all the content in the spec up till page 18, use “fault”, including example, but from section 4 onward, all the faults become “failure”, for
consistency’s sake, and a common sense in SOAP, “fault” is better here.
RNS properties have been described twice in the spec, one in section 2.2, one in section 3, don’t think it is necessary, or can make section 3 a sub
section in section 2.2.
the <entry-name>rns:EntryNameType seems to be rather important, and serving as kind of key for all the operations, but apart from the restrictions on
its style, do we also need to consider its uniqueness?
That is all I can think of right now, if it is too late for public comment, I would like to share this questions with the authors as well.
Thanks,
Vivian
Dave,
After reading “RNS-OGSA-rendering”, I sort of see why they name it “RNSResponseEntry” rather than “RNSEntryResponse”, but still, sounds odd. The
rendering document provides nothing more than schema and wsdl, so not much comment there. However, it does reflect the question about using the word
“failure”, my personal opinion is to use fault throughout, just to avoid confusion. Another finding is back to the spec, all the namespaces of s11
in the example is inconsistent to section 1.3, not that it is important, but they should be consistent.
Thanks,
Vivian
|
