Hi,
(feel free to drop names and lists if you think the audience is too
wide)
RFC 2119 and Conformance is a whole another topic… which has created
discussions in the belated QA WG.
Le 16 janv. 2007 à 03:49, Dan Connolly a écrit :
Looking at [1], I see
"An Assertion must have at least the following properties"
that's odd too.
[[
I think it's a misuse of RFC2119 to say things like "2 + 2 MUST be
4" or "every attribute value in an XML document MUST be quoted."
Better to just say "2 + 2 is 4" and "every attribute value in an
XML document is quoted."
]]
-- "must is for agents", Dan Connolly, Jan 2001
http://www.w3.org/2001/01/mp23
See also
http://esw.w3.org/topic/RfcKeywords
With Danc's proposal here, let's state that we only use "RFC 2119"
for agents. There is then a need sometimes to define in a
specification what is an assertion and what is simple prose. The only
way to do that is, either, having
- separate list of testable assertion
http://www.w3.org/TR/qaframe-spec/#write-assertion-gp
http://www.w3.org/TR/qaframe-spec/#consistent-style-principle
- a specific markup in the specification to identify what is prose
QA Specification Guidelines says:
http://www.w3.org/TR/qaframe-spec/#norm-informative-gp
"Specify in the conformance clause how to distinguish
normative from informative content."
with the associated technique
"3. Try to avoid language that sounds normative in an
informative section. It might lead the readers to wrong
assumptions."
But sometimes it is not that easy. Forbidding the use of RFC 2119
keywords is odd too, there are just English terms.
Question:
- Does the term "agents" include humans?
--
Karl Dubost - http://www.w3.org/People/karl/
W3C Conformance Manager, QA Activity Lead
QA Weblog - http://www.w3.org/QA/
*** Be Strict To Be Cool ***
