Draft minutes: Crete F2F - June 16, afternoon

June 16, afternoon
Scribe: Patrick Curran


New AIs

Dom & Lynn: estimate the dates for Spec GL
   Due: before we leave Crete
Daniel: create outline of document presenting arguments for value of "QA 
now"
   Due: 1 month
Daniel: draft new WG and IG charters
  Due: 1 month

1. New EO project(s) -- cost-now vs. cost-later for QA

QA cost now is cheaper than later

Daniel: should we write a paper explaining this?
Who to target: the WGs or the members (who would fund it)?
Would be good to target the members - need a business-oriented argument

Dom: might be difficult to persuade people that this is true
Patrick: the argument works inside Sun

Daniel/Karl: convince people that the "cost" later is lack of 
interoperability
Patrick: the saving is in test development (share the burden)

Daniel: do we need make arguments re compliance with 'monopolists'?
Karl: some people don't care - eg, "we only need to develop for IE"
Daniel: agrees

Mark: provide good test suites and people will use them

Daniel: anyone want to volunteer to write this paper? Business justification
Two different cases: addressed to a monopolist and a non-monopolist

Dom: how would we ensure that the document reached the right people?
Daniel: potential WG members could use this to convince their own management

Patrick: repeated the arguments he used to convince Sun to participate
  We will get out more than we put in
Lynn: there can be a competitive element to this:
  if vendor 'a' submits, vendor 'b' might follow suit
  quote a success story

Daniel: the financial argument (pick up these free test suites) is 
secondary;
how do we convince people that compliance is important?

Karl: convince the consumers (example: need for well-structured documents)
Daniel: couldn't convince his daughters
Patrick: don't try to convince the end-users, but an IT head will be 
easier to convince
Mark: user-demand, or government regulation will convince people
  W3C has been successful in convincing people that standards are good

AI: Daniel to create outline of this document (due in 1 month)
Karl: post the outline, encourage discussion on mailing list


2. List policies / logistics review

Lofton: do we need an anti-spam mechanism? Respond by email to submitter.
Dom: explains how it works
We don't actually have a problem - this is moderated
No need to change current policy


3. Other IG and outreach topics TBD

Olivier: briefly lists documents they are working on for the QA library
Daniel: how do we decide what to work on, how to prioritize?


4. The Framework/GL specs, per GL (Ops/Spec/Test)
  * status update
  * changes to goals? (Rec track, etc)
  * immediate next steps
  * medium-range plan/schedule

Lofton: we should try to get the docs to "good enough" status asap,
so we can move on to other stuff.

Status:

Lofton: Intro is in decent shape - only a few issues. Since not normative,
this should probably be converted to a Note.

[ = ET ]
                    Ops GL/ET        Spec GL/ET        Test GL/ET
Public Working Draft            5th            4th            2nd
(LC) issues                0 (all resolved)        90-95% resolved    no 
issues
Revised text (after issues)        75% [0%]        60-70% [0%]        N/A
Verify text (ed draft)                                    N/A
Review revised text
Disposition of Comment responses    Publish WD        Publish WD
Negotiate DoCs with submitters
Next publication (what state?)        CR+ [ET done]    CR+ [ET done]
Final goal?                TBD            TBD

Ops: changes since last draft are relatively minor. No need to publish 
another Last Call.
Republish another working draft for convenience of readers. Then we can 
go to CR
(Candidate Recommendation)

Spec: discussion whether we should go to another Last Call. Lofton - 
other Chairs
believe that if there are significant changes we should not go to last 
call. Dom believes
that we are still substantially the same document - can go to CR.

Daniel: we could do a special CR (leave it in this state for several 
months).

Discussion: should we convert to Note?

Lofton: we could simply ask WGs to fill out checklists (while our docs 
are in this
extended CR state) - there would be no need to conform.
Karl: wonders what the benefits would be
Dom: this would enable us to gather feedback (we'd have to ask for it).
Karl: we would get an idea of how well we're doing
Lofton: we didn't get a lot of feedback
Dom/Daniel: we did pretty well - don't downplay it
Daniel: make the checklists good quality - should point to real-world 
examples
Lofton: this means we would have to improve Examples & Techniques

Discussion: should we try to keep Ops and Spec docs in sync?
Not necessarily.

Fill in the dates. Lofton - believes conservatively he can reach the LC 
state
within 6-8 weeks. Spec will probably take 4-6 weeks more.

AI: Dom & Lynn to estimate the dates for Spec GL

Daniel: duration of the Special CR? at least 6 months. Synchronize the 
ending
period (for next year's Tech Forum)?

Lofton: what about Test GL? Should we go straight to Last Call?
Dom: don't go to Last Call just because this is the only way to get external
comments. If we're not ready, don't do it.
Goal: LC after October meeting. Can we do this, while people will be working
on SpecGL? Needs an ExTech before we can publish.
Lofton: should we suspend it until we have time?
Continue the discussion after we've covered this doc during the next
couple of days.
Patrick to take lead editor role? He will consider.
Lofton: we shouldn't publish another version before LC.

Lofton: is Recommendation where we want to end up? (Daniel has argued
that this may not be appropriate - Note may be better.)
Dom: mildly agrees.


4. Expiration of QAEG and QAIG charters
  * recharter? [assume "yes"...]
  * proposed Statement Of Work for next 1-2 years
  * staff & member resources for new SOW
  * drafting a new charter
    * when needed? (for final draft?)
    * who? (for initial draft?)
    * how? (QAWG review cycle?)
    * when initial draft?

What to do once we get the docs completed?
Suggestions: docs, templates, outreach, feedback (publish best practices
that have been adopted by others), review of other WGs' docs.

Reviewing the current charter's deliverables

* maintainance of an up-to-date QAWG Web site (/QA/WG/)
  OK
* timely minutes of teleconferences and face-to-face meetings
  OK
* up-to-date matrix of specifications/QA activities (a.k.a. The Matrix 
(../TheMatrix.html));
  OK
* glossary of QA terms and taxonomy (../Taxonomy.html) used in W3C;
  OK
* improved editor guidelines for specification writers (aka pubrules++ 
(/2001/01/qa-ws/pp/susan-lesch-w3c.html), including conformance section, 
list of testable statements, list of explicit discretionary behaviors, 
etc...)
  This is SpecGL - new charter should state that we will continue to 
work on/improve Examples & Tools
* report on QA review of specification conducted
  Rewrite as ongoing proactive reviews of other WGs' specs
  Or - do we want to create a summary report?
  Can we create a feedback loop? How to capture best practices?
  We are committing to ongoing reviews, and to working to improve ExTech 
docs
  make it clear that these are connected
* issue resolution related to appeal/vagueness/interpretation of specs
  Delete this - the WGs must define their own processes
* proposed changes to W3C Process Document (/Consortium/Process) and 
standardization cycle (from early draft to Candidate Rec, going to the 
AB, as the group responsible for the evolution of the W3C process)
  Rephrase this to refer to OpsGL (parallel to the SpecGL rewrite)
* report of review of QA tools conducted in the activity (WG or IG)
  This is similar to the stuff in the TTF charter
  Should we have a separate task force?
  Think carefully about how to phrase and balance/prioritize to bring in 
new blood
* a framework/process for developing, managing and running tests (data 
driven, platform independent, formal
output language for results - e.g. EARL (/2001/03/earl/), etc)
  This is Test GL/ExTech
* Promoting existing Notes on QA as W3C Notes and developing new ones
  An IG activity?
  An ongoing activity - no need to call this out explicitly as a deliverable


Proposed TTF Charter

Which of these general deliverables from TTF charter should we move to 
the new charter?

* Provide information on existing Test Technologies and techniques used 
and developed by Working Groups in the W3C and other interested parties 
(information can be documents, tutorials, web sites and so forth)
* Provide on-demand expertise to Working Groups seeking to produce such 
materials
* Assure uniformity in Test Frameworks and Techniques produced, as 
explained in other QA WG documents (ensure that tools conforming to QA 
WG Test Guidelines documents are used and, where applicable, reused by 
as many WG as possible).
* Develop tools and tool kits of general usefulness to help WGs develop 
test materials, resources allowing.

What to do with the detailed "for example" list of potential 
deliverables from TTF charter? Link
from WG charter?

AI: Daniel will draft new WG and IG charters within one month

Received on Wednesday, 18 June 2003 07:05:19 UTC