W3C home > Mailing lists > Public > www-qa-wg@w3.org > March 2005

TAG issues -- enumeration and opinions

From: Lofton Henderson <lofton@rockynet.com>
Date: Thu, 03 Mar 2005 10:05:27 -0700
Message-Id: <>
To: www-qa-wg@w3.org


I have gone through the TAG comments.  Here I index and summarize them.  I 
also give my opinions.

Some will probably (IMO) want at least a little WG discussion, even tho' we 
probably agree in principle with most of 'em (i.e., they will probably be 
pretty easy to resolve):
1, 2, 3, 5, 9, 10, 13, 14, 16, 17

Some are (IMO) editorial or completely straightforward:
4, 6, 7, 8, 11, 12, 15


Overall comment, requires no response
Praises document as good, easy to read, useful, etc.

*TAG.1: positive statements
Suggests positive statements, "No deprecated features."  (Applicable to 
stuff like deprecation, extension, profiling, etc.) [LH:  I agree.  This 
has been discussed before.]

*TAG.2:  workflow
Asserts that we mix specifications and process (workflow) -- they should be 
distinguished, and questions dealing with them "in the same 
proforma".  [LH:  No specifics given, I'm unsure.]

*TAG.3:  optionality of conformance
About putting "no conformance to this document" elsewhere than in a 
Conformance Clause, suggests template wording for that case.  [LH:  I think 
it should be recommended to be in a CC, not elsewhere.  This has been 
discussed before.]

Wants TOC to go down to level 3 (listing GPs and RQs).  [LH:  I agree, at 
least that they should be enumerated with links in or near the TOC.]

*TAG.5:  "then Level 1 too"
Questions 4th bullet of figure 3.  [LH:  I agree it's confusing and 
possibly wrong.  What if Profile X placed restriction, 
'foo-level="2"'?  Does that even make sense?  At least clarify the bullet 
to answer TAG questions.]

TAG.6:  URI and normative refs
a.) grammatical critique of language in Examples of GP2.3B [LH: agree]
b.) suggests potentially more discussion in 3rd & 5th  [LH: no opinion]

TAG.7:  normative refs, Unicode and XML
Suggests referencing Unicode and XML (1.0, 1.1, ...) as common 
examples.  [LH: I agree.]

TAG.8:  warn implementors
Suggests improved language [editorial] and previews next.  [LH:  agree, 

*TAG.9:  warn implementors (2nd)
Asserts that "Implementors" are not the only ones who extend.  [LH:  I 
agree with the assertion.]

*TAG.10:  deprecated features
a.)  asserts that even v1 can deprecate.  [LH:  unsure about this.]
b.)  if no deprecation, make positive statement "No deprecated 
features."  [LH:  I agree.  See also TAG.1]

TAG.11:  define deprecated features
Suggests more precise rewording [editorial?].  [LH:  I agree.]

TAG.12:  which of (editorial)
[LH:  yes, editorial.]

*TAG.13:  neither language nor protocol
Points out that we only prescribe error treatment for languages and 
protocols, suggests there is other eligible stuff.  [LH:  I agree.]

*TAG.14:  use your own examples
The story in Section 5 is about SpecGL CR, and should say so.

[LH:  IMO, yes, but as I discussed with Dom, the Story is partly 
fictionalized.  I believe that not all of the problems were due to lack of 
quality review, but were conscious decisions of QAWG -- e.g., that ExTech 
could be maintained in WG space and developed more slowly.  JC didn't 
agree.  It's a good story, but if we take credit for it, then we ought to 
make it more accurate.]

TAG.15:  edition (editorial)
[LH:  yes, editorial.]

*TAG.16:  specgl-ics
The ICS should have not just y/n/na, but the possibility to link answers to 
tests and other stuff that supports the answer.  [LH:  I agree.  This is 
the same as Ian Hickson's comment, bug #1041.]

*TAG.17:  conformance to self
He applied SpecGL ICS to SpecGL.  It has some NO answers and should 
not.  (Most failures are related to other issues in the TAG 
comments.)  [LH:  I agree.]
Received on Thursday, 3 March 2005 17:05:39 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:43:38 UTC