Final Minutes for 2002-07-31 telcon

  QA Working Group Teleconference Final Minutes
Wednesday, 31-July-2002
Scribe: [Andrew Thackrah]

(DD) Dimitris Dimitriadis (Ontologicon)
(DH) Dominique Hazael-Massieux (W3C - Webmaster)
(KD) Karl Dubost (W3C, WG co-chair)
(DH) Dominique Hazael-Massieux (W3C - Webmaster)
(LH) Lofton Henderson (CGMO - WG co-chair)
(SM) Sandra Martinez (NIST)
(JM) Jack Morrison (Sun)
(AT) Andrew Thackrah (Open Group)

(LR) Lynne Rosenthal (NIST - IG co-chair)
(KG) Kirill Gavrylyuk (Microsoft)
(MS) Mark Skall (NIST)

Absent: (PF) Peter Fawcett (RealNetworks)

Summary of New Action Items:

A-2002-07-31-1: DH to investigate web requirements (eg CGI) to handle 
questionaire form

A-2002-07-31-2: KD to send feedback request message to team contacts

A-2002-07-31-3: LH to send feedback request message to chairs list

A-2002-07-31-4: KD to choose the review subject for any QAWG members who 
have not already chosen a subject

A-2002-07-31-5: LH to ask David Marston if he would like to carry out a 

A-2002-07-31-6:  DD to complete questionnaire final draft by 2002-08-14 
and circulate to QAWG


Previous Telcon Minutes:

2) Logistical topics

LH: About the Tokyo face-to-face meeting. See agenda link [0] for 
preliminary     information. About the hotel: it is clean and cheap. The 
meeting is     definite. Airfares have dropped in the last month by 30-40%

KD: The hotel is very good. It is convenient for the train and practical 
for         exploring the city.

3) doc-tech questionnaire

LH: Regarding the call for comments on the questionnaire ( agenda link 
[1]). There have been     no further comments. I think we should not wait 
for further comments. We should     go ahead and use it now. It's 
important to get this in use - it is a prerequisite     for our goal of 
prototyping a test assertion markup language.

     It needs an introduction.

DD: yes, it does need an intro.

LH: Apart from the lack of an introduction, one thing is missing: how do 
we get feedback?
     Should we use team contacts? (they could deliver to WG editors for 
dispersal among members).     Or do we invite feedback direct to the QAWG 
list? Or direct to Dimitris for further editing?

KD: In Montreal we agreed to make a web-based form for people to submit 

DH: We need the cgi scrips to process this. There maybe some materials 
available already.
     [Action Item for DH to investigate this]

DD: Can we make this (quasi)mandatory? Chair AND team contacts should be 

KD: agreed - ask both chair and team contacts.

DD: The response should only take ~5 minutes to complete

[Action Item for Karl to approach team contacts]
[Action Item for Lofton to approach group chairs list]

DD: Shall we set the due date for final text as August 15?

DH: This date is ok for me

DD: A one month deadline for groups to respond after we publish should be 
     We should use strong wording to encourage a response.

LH: Agreed, though we can't mandate it because members are volunteers.

4) Review Assignments

DD: I have to leave now. Please assign a review for me.

LH: review matrix is more empty than full despite calls for volunteers. 
maybe we should     volunteer people to fill the gaps?

KD: Can we have a roll call now to see who has chosen yet?

LH: Karl has some blanks
     Dom has some blanks
     What about David? should he be included?

KD: Can we select different groups for each review?

LH: Yes, but it may help to use the same groups

DH: Should we review ourselves?

LH: Good idea

DH: OK I'll volunteer to review QA

KD: The person doing the review should understand the topic of the spec if 

LH: Yes. For me CSS was harder than SVG because I had less knowledge

KD: I propose that we ask people to choose by next week. And if they don't 
then     a group is selected for them.

DH: In that case we should give the person a week to appeal

SM: What is the 'tdb' in the review matrix?

LH: 'tbd' represents the date for completion. You must pick a date as well 
as a group.     But we don't want all dates to coincide. We want to 
stagger them over major document     versions.     But I would like some 
feedback on the current version - in the next two weeks.

SM: Will send URL for XML

DH: Andrew has chosen HTML but there are many documents within that area - 
what should he do?

AT: I'm interested in modularization

KD: I recommend a joint review of Modularization and XHTML 1.1

AT: OK, I'll do that

[Action Item for KD to choose review subjects for people who have not yet 

[Action Item for LH to ask David if he wants to participate in the review 

LH: The review matrix needs a 4th column with title 'Techniques assessment 
for a GL document'

DH: Done - looks OK

JM: Can we link the completed reviews - this would be helpful

LH: OK, but they still need cleaning up before publishing so I can't do it 

5) Spec Guidelines document

LH: In Montreal we decided that we would publish an interim version in 
August     (before the major October publication date) because of the 
extra interest and     resulting reorganisation of the document. We should 
aim to publish in week 2 or     3 of August so we must resolve the content 
in the next two weeks. We have one regular     telcon and if necessary we 
can have an extra telcon.     So we have a deadline to finalize majpr 
content details by (say) August 14.

     We have one major thing missing: a checkpoint-by-checkpoint priority 
review     So next week we have to complete/endorse priorities

     Then, after publication Lynne will take back ownership.

AT: Is the deadline realistic?

LH: what do people think?

AT: Yes with the caveat that we sort out the distinction between 
levels,     modules and profiles.  I'm confused by this and I think spec 
authors will be too.

LH: Yes, we have inherited a choatic legacy from the W3C specs. But 
it's     starting to converge. We have to address the reality of what W3C 
has produced     and what our ideal is.

SM: I agree with Andrew. Reviews can help here also. We need 1 or 2 more 
reviews for this     version. [SM leaves]

LH: This does seem to be a significant issue

KD: It would be better if we have more input from editors

LH: Should we postpone your (Karl's) issues discussion today?

KD: let's put it to the end of the agenda

LH: So is SpecGL close enough given that we only have 2 weeks?

KD: It has good features but sometimes as an editor who will have to use 
it I     think that it will be difficult to implement. We should have a 
manual!     As an editor I may not be sure whether I am supposed to use a 
feature or not.

JM: Based on past experience I think we will need that extra telcon

LH: Yes, probably. We could manage 3 telcons in August: 6th, 13th, 21st? - 
that     leaves 10 days in the month for editing.

KD: Should ExTech be published at the same time? This may be difficult 
because     it requires a stable SpecGL.

LH: No problem there, ExTEch can wait.

KD: yes - wait for SpecGL stablity

[Discussion of KD's SpecGL comments follows]

KD: #1 Define what we mean by a table of content entry - where does it 
lie?, what form? etc
       Sometimes it if not clear if text should be included in the ToC. 
Should work with       Susan Lesch on this.

LH: Have you seen the new draft - I tried to reflect last week's comments 
about     that in checkpoint 3.2, 3.4 etc - by adding a 'not applicable' 
     A spec must say something about profiles for example so that should 
always have an entry

    The actual wording tries to address Karl's issue. Karl - is this 
sufficient for the next    public draft?

KD: Seems fine - but in XHTML for example sometimes the ToC does not 
contain all links. It's     possible to have an entry on one page that is 
not in the ToC.

    [looking at XHTML 1.0, comparing sections with ToC - this example 
challenges the     ToC definition]

LH: Given the new text do we need to go further?

KD: It's better

LH: Look at (SpecGL 2002/07/29) GL 8, ckpt 8.5 - are these detailed enough 
for the next draft?

KD: I will it review it later today - seems to be ok

[ LH invites everyone to see if the vatious ToC-related checkpoints are 
suitable for publication    in the next draft version ]

KD: #2 Clarify modules and levels

LH: Sandra and Andrew also flagged this - we should deal with this by email

KD: #3 ?? [KD comment not recorded]

LH: In GL 3, 4 and 7 - have put a 'not applicable' qualifier (except on 
ToC checkpoints) 
KD: Fine, this addresses my comment

KD: #4 When people start a new technology - I don't know if they will be 
considering     levels at the start.

LH: Checkpoint 7.1 tries to capture this


LH: See David Marton's email about mods and levels

     One case is a high precision level and low precision level designed 
in from the     start. Don't think we have an example of this in W3C?

KD: What, a first edition of a standard that defines levels? Not sure.

DH: DOM level 1 - this appeared just before PR version, but not in the 
first working draft

LH: CSS didn't predict levels either.     So, is Checkpoint 7.1 ok for now

KD: I will review it and confirm

LH: Generally, we have to clarify levels/modules/profiles
         Next meeting is next Wednesday, on the regular schedule. We can 
also expect an extra     telcon to discuss the current draft of SpecGL



Received on Wednesday, 7 August 2002 10:15:42 UTC