Minutes from the QA WG 2002-08-21 teleconference

QA Working Group Teleconference
Wednesday, 21-August-2002
--
Scribe: Lynne

Attendees:
(dd) Dimitris Dimitriadis (Ontologicon)
(KD) Karl Dubost (W3C, WG co-chair)
(DH) Dominique Hazael-Massieux (W3C - Webmaster)
(LH) Lofton Henderson (CGMO - WG co-chair)
(DM) David Marston (IBM)
(SM) Sandra Martinez (NIST)
(LR) Lynne Rosenthal (NIST - IG co-chair)

Regrets:
(KG) Kirill Gavrylyuk (Microsoft)
(MS) Mark Skall (NIST)
(AT) Andrew Thackrah (Open Group)
(JM) Jack Morrison (Sun)

Absent:
(PF) Peter Fawcett (RealNetworks)

Summary of New Action Items:
-AI-20020821-1 dd to construct email message for sending out questionnaire 
[1].  Due Monday 26 Aug. 
http://lists.w3.org/Archives/Public/www-qa-wg/2002Aug/0132.html
-AI-20020821-2 LH to send questionnaire to chairs list. Due Monday 2 Sept
-AI-20020821-3 DH to send questionnaire to Team. Due Monday 2 Sept
-AI-20020821-4 LH to write up an issue for CK10-5  handling of legacy 
documents
-AI-20020821-5 LH to write up an issue for GL 8  should there be a new CK 
to require explicit documentation of interoperability impacts
-AI-20020821-6 LH to write ground rules for techniques analysis review and 
call for sign ups.

Agenda:  http://lists.w3.org/Archives/Public/www-qa-wg/2002Aug/0134.html
Previous Telcon Minutes: 
http://lists.w3.org/Archives/Public/www-qa/2002Aug/0055.html

Minutes:
1. Roll call
See above

2. Logistical Topics
*F2F Meeting in Tokyo. Details given at http://www.w3.org/QA/2002/10/f2f.
If you haven't done so yet, please make your hotel reservations.
Plans are to provide wireless network connection.  If you prefer a wired 
network connection, contact Olivier directly and cc the qa-wg

[Since dd leave early, do Agenda item 4 prior to 3]

4. Questionnaire for doc technologies
http://lists.w3.org/Archives/Public/www-qa-wg/2002Aug/0132.html

*Questionnaire completion
dd sent draft of questionnaire to WG list earlier today.  An introduction 
was added as well as a clarification question as proposed by LH.  The 
introduction provides a rationale for using structured markup, indicates 
that we look forward to receiving comments, and that it will only take 
about 5 minutes to complete.
ACTION for All: Anyone who wants to supply feedback on the questionnaire 
has until 24 August (the end of this week).  See AI-20020821-1 (for dd)

*Interactive web form (CGI form) or as email message.
In an email, Olivier identified a form from another WG as a possible 
starting point  from which we could adapt for our use.  DH has concerns 
about the maintenance and security of using the form.  dd - It would be 
necessary for someone to adapt the form and rewriting may take longer than 
to write one from scratch. DH, depends on how it was done  maybe, maybe 
not.  Do we want to spend time on an interactive form? LH  would we be more 
likely to get people to fill the questionnaire if we had an interactive 
form vs. email with checkboxes?  DH may be easier to have as an email, but 
if interactive form, we can do queries.  Using CGI script would provide for 
simplicity in processing the results. LH wants the questionnaire to go out 
week after next, so we have results by F2F.  DH recommends email, since 
only expect about 20 replies (since there are 30 WGs)  Emails would be 
archived.  DH asked if there would be problems with member 
confidentiality?  dd doesn't think so since the questionnaire doesn't ask 
those types of questions. DH suggests using email and having people respond 
to a specific person (e.g., dd or DH) as well as to the 
qa-wg.  AGREED  send questionnaire via email.  See AI-20020821-2 and 
AI-20020821-3 (for LH and DH)

3. Action Items Review/cleanup
http://www.w3.org/QA/WG/ActionItems
*Since most people who have AI's are not on this call, will postpone going 
through the list. Only address dd's AIs.

*Status of dd's AIs
A-20020731 Questionnaire will be finished by Monday 26
A-20020814 Done.  Sent text to editor, clarifying the wording of 
checkpoints for GL 14 and 15
A-20020614 Not yet started to develop a proposal for forming a testing activity
A-20020614 Waiting for results of questionnaire for terms for testable 
assertions

[dd leave call]

5, Spec Guidelines  [5]

*No comments regarding changed needed for SpecGL release on 26 August.

*LH had a few specific issues to discuss
http://lists.w3.org/Archives/Public/www-qa-wg/2002Aug/0115.html

*Status section contains a list of the kind of attention and feedback we 
are soliciting.  LH- is this a good summary of what we want?  All agreed it 
was.

*There are still some issues regarding atomicity, levels-modules, etc. LH 
-What issues do we want to call attention to?  In status section, we could 
reference the issues list, including specific issue numbers or just provide 
the link to the issues list.  KD suggests that it's better to link to the 
specific issues.  LH will update the status section by 23 August.

*Goodness criteria for dimensions.  DM suggests the document address the 
goodness criteria and that the overall number of dimensions can cause a 
breakdown in interoperability. LH indicated that the document hints at 
this, but it probably is good to make it an issue.  DM including a formula 
such as numbers of dimensions or variability score, could stir up a 
reaction.  LH will add this to the issues list and also flag it in the 
status section.

*CK10.5  the proposed new text removes the negative disclaimer.  The new 
text was DM's assignment and implemented by LH.  It resulted in a new 
checkpoint, CK10.5 and adjustments to related ckpoints (e.g., 3.1).  Is the 
new text acceptable?  No comments.   DM pointed out that this is the only 
place in the document that uses the date of the specification as a dividing 
line.  The objective is to allow legacy documents to meet the spirit of the 
GL and not be penalized for not explicitly calling out that there are no 
levels, or profiles or using the terminology of the SpecGL.  DH asked why 
we do this since it is unlikely that older specs would be evaluated (rated) 
against the GL.  LH  it was decided (last telcon) to phrase things so that 
legacy documents that have observed good practices in the spirit of the GL 
could get a good score.  Should this be a new issue? DH  make it a priority 
so that it is addressed soon.  DM  when this draft is published, other WG's 
could start to use the terminology so that their document progress in 
parallel with the QA documents. The objective is to get some convergence to 
occur prior to the final publication of the SpecGL.  LH  The Intro document 
will eventually be amended to describe the mandatory-ness of these 
documents. Meanwhile, in the status section, we can highlight what DM 
said  that is, we want people to start looking at concepts and terms now 
and start to use them; migrating towards them now.  LH will include 
something in the status section regarding the mandatory nature of this 
document and a suggestion to start using it now. DH asked if we should do 
this at all  that is, pay attention to legacy documents.  LH will put this 
on the issues list.

* As a result of 10.5, LH amended 3.1: deleted last sentence and provided 
new text.  How is a legacy document able to pass? DM suggested adding more 
verbiage, like an example that says that for legacy document to get a 
passing score, it should be clear to a reader of the document that there 
are no profiles (even if they don't say "there are no profiles" 
explicitly). If it isn't clear that the legacy document would fail the 
checkpoint. LH asked if anyone had a better idea.  No, then, rather leave 
it vague now or say explicitly that the checkpont is NA and not fail.   It 
was decided to leave it as is (i.e., vague for now), even though it is 
ambiguous as to how to do the score sheet.

*GL8- discretionary.  Do we want/need a new checkpoint?  Should there be a 
new checkpoint requiring that interoperability effects of dependencies be 
documented  i.e., requiring explicit documentation of interoperability 
impacts.  No resolution at this time.  Still an open issue.

6 Next meetings/projects

* 3 telcon scheduled between now and F2F.  (4 Sept, 18 Sept, and 2 Oct)
LH - with the publication of SpecGL, we have met a major milestone.  Once 
its published we can expect to get comments and that will help to define 
our program going forward.  If we don't get a lot of comments, we have more 
flexibility on moving forward and can discuss this during the next 
telcon.  The F2F will focus on the publication of all document parts for 
late October.

*The focus of the next telcon will be on Issues list and AI lists. If 
Kirill publishes the draft TestGL before the end of August, the telcon will 
be dedicated to looking at the draft.  As for the Spec Examples and 
Techniques document, KD wants to create a new version prior to a 
discussion.  He will probably have a draft by late September and it can be 
discussed during the 4 October telcon.

*Matrix Reviews.  LH - Although the matrix is populated, there still needs 
to be dates added indicating when we want people to do reviews over the 
remainder of year.  Want to have some Spec GL reviews with new draft and 
also after Oct with publication.  There is a blank column for people to do 
techniques assessment of a document.  Also, want to divide people up into 
groups of three.  As soon as a cell has three people, it is closed.  At the 
next telcon,

7. Issues list items
Postponed, too late to start now.

Next telcon 4 Sept.

Adjourn
11:35