Re: [office-comment] ODF 1.2 References

[marbux <marbux@gmail.com>]

On Mon, Feb 23, 2009 at 10:21 AM, Dennis E. Hamilton
<dennis.hamilton@acm.org> wrote:

> That textual examination of the use of terms found in conformance language is very useful.
>
> I want to point out that, for the ISO/IEC nomenclature, MAY and NEED NOT go together

Yes, I erred by omitting discussion of NEED NOT and of CAN and CAN
NOT, each of which is defined by ISO/IEC Directives Part 2.

and MAY NOT (as a prohibition) is not acceptable normative language.

Lawyers and judges love to use the term in that sense, but doing so
departs from common usage and if not used carefully in context can
leave ambiguity. I agree that avoiding the term is best because the
words in combination are not a defined requirement term in the
Directives.

> Also, I think sometimes the use of MAY and MAY NOT is more in the sense of CAN and CAN NOT (with regard to possibility, as opposed to permission or prohibition).

Agreed.

> I have also noted that SHOULD in the ISO/IEC sense is quite different, and considerably weaker as a prescription, from the SHOULD as used in the IETF sense.

Agreed. I wonder if it would be useful to include an informational
note to section 1.2 calling attention to the fact that ISO/IEC
Directives definitions differ in substance from the RFC 2119
definitions used in OASIS ODF v. 1.0. Nearly all XML-based standards
I've looked at use RFC 2119 definitions, so ODF and a few other XML
specs that have been adopted as international standards are outliers
in this regard for the developer community writing XML
implementations.

> (You will notice how I indicate bold face in plaintext, which is probably the same reason that the IETF uses capitalization, since the authoritative texts of IETF RFCs are in plaintext.)

Yes, but note that W3C aligns with IETF in this regard and W3C
specifications are published in HTML, which would allow use other
means of emphasis such as text attributes. The OASIS guidelines for
drafting conformance clauses explicitly calls for RFC 2119 terms to be
emphasized with all-capital characters, but is silent on the subject
of emphasis for ISO/IEC terms.

> (I also assume that technical use of a style in the formatting of the ODF version is not that helpful for the PDF version of the document.)

> On the other hand, one potential assist in the clarification of when normative terms are being used is from the OASIS guidelines for conformance sections and conformance language.  There, there is a notion that every statement that has conformance language in it is distinguished in some way (the W3C does this, for example), there is some form of index or compilation of all of them, and each of them can be traced to a conformance target specified in the Conformance section of the specification.

Conceivably, this could be the editor's purpose in marking such terms
with a custom style, which would facilitate automated extraction of
paragraphs containing the terms.

Best regards,

Paul


-- 
Universal Interoperability Council
<http:www.universal-interop-council.org>