(unknown charset) Draft minutes of QAWG Telcon 2004-07-26

DRAFT

QA Working Group Teleconference
Monday, 26-July-2004
--
Scribe: Andrew Thackrah (Open Group)

Attendees:
(PC) Patrick Curran (Sun Microsystems)
(DD) Dimitris Dimitriadis (Ontologicon)
(KD) Karl Dubost (W3C, WG co-chair)
(DH) Dominique Hazaël-Massieux (W3C)
(LH) Lofton Henderson (CGMO - WG co-chair)
(LR) Lynne Rosenthal (NIST - IG co-chair)
(MS) Mark Skall (NIST)
(AT) Andrew Thackrah (Open Group)

Guest:
(DM) David Marston (IBM Research)

Summary of New Action Items:

 LH To rewrite the TestGL document status section as agreed. Due 2004-08-09 
 DH to send email to TAG to explain how we are looking at the issue. Due 2004-08-09

Agenda: http://lists.w3.org/Archives/Public/www-qa-wg/2004Jul/0053.html 
Previous Telcon Minutes: http://lists.w3.org/Archives/Public/www-qa-wg/2004Jul/0046.html 

Minutes:
 1.) roll call 11am EDT, membership

 2.) routine business

 - Future telecons   

 (Monday 9 August): DD & DH on holiday for next telcon
				
 Mark Skall and NIST: Changes in organization.

 MS: 3 weeks ago I was made acting director of ITL, so it may be 
 difficult for me to participate.

 - Texts from NY Panel organized by Dave Marston
		
 KD: Could we put together the texts presented at NY? 
 DM: There was no recording going on at meeting
 MS: I can send my notes
 DD: I have no minutes/notes 
 KD: could you provide 1-2 paras of explanation
 DD: sure
 KD: please send it all to DM

 [ Please send all notes & minutes of presentations to Dave ]
		
 KD: Extra item - a question from LH - about ExTech - cf the email link [1] - 
 this doesn't reflect current thinking about test-lite. So maybe put a place-holder 
 there.  Use a pointer to previous version & point to new ideas/wiki etc

 DH: It's possible & recommended to do this (i.e. to give regular updates).

 DD: Please ensure that we say 'further publication has been postponed'

 LH: I'm concerned that status section will be overlooked and people will think that 
 the February draft is latest stage of our work. So prefer to not link to old draft

 KD: We could make changes obvious

 DH: We could remove everything but placeholder and detail the new direction

 LH: Previous version would point to Feb., status section  would ref. new resources

 KD: So delete the body of the doc. - keep status
		
 [ AI LH To rewrite the TestGL document status section as agreed. due 2 weeks ]
		
 Aside: LH: I tried to reference a wiki item. I looked at source and found 
 a name atrribute but no id attribute. Isn't name attribute deprecated?

 DH: Wiki s/w to blame. We tried to contribute to s/w but developers are 
 protective
		 

  3.) How to manage normative references in W3C Specifications
        Issue Björn Hörhmann [1]
        Dicussion Lead: Karl [2]

 KD: Implication of pointing to a ref'd spec. includes a feature taken from 
	another spec. So ref. has to be precise.

 DD: We should try to get a very basic solution.

 DM: Everytime you cite a spec - have to be clear if you mean an exact version 
 or if you want your citation to track the spec.

 KD: Tracking a spec could lead to problems

 DM: E.g. XML 1.1 names & Unicode has this tracking.

 LH: It's a fairly rare circumstance - what shoud spec-lite attitude be - treat 
 it as a rarity?

 KD: I'm not comfortable with it - we can't predict future of a spec - 
 e.g. deprecated features

 DM: One could imagine it with functions & operators (for example reusing XPATH material)

 PC: Examples where a ref. is strictly normative are very rare and dangerous 
 [@@ AT - did I hear this right?] . Maybe 
 we should encourage but not mandate tracking to future specs

 DH: There's the relationship with extensibility to consider

 KD: Yes, where should it go in the doc?
	
 DH: Maybe section B.3 & rename it to be more general
 
 [ agreement that discussion should go in B.3]	

  4.) XML and Extensibility
        Discussion Lead: Dom [3]
        Dave Marston has a suggestion [4]
        """
        Would it be enough for SpecGL to say that every
        WG must assume that there will be a newer version
        of every spec? Maybe we could add in some principles
        that are already well established, such as: everything
        in the XML world must be marked with a version
        identifier.
        """
	
 DM: Both the article & Dave Orchard article recognise that versioning gets complicated 
 so propose use extension mechansims to solve 80-90% of these issues.

 MS: Version - this may be an Ops issue, rather than Spec issue?

 DH: Would someone like to volunteer to review the artcle?

 DM: I did analyse the other article. TAG is new - everyone saying the we can't solve the 
 entire thing but TAG will eventually produce something like an OpsGL. So SpecGL could 
 say 'use the TAG recommendation'
		
 KD: So what overlap do we have with the web arch doc?
		
 [ AI - DH to send email to TAG to explain how we are looking at the issue. due 2 weeks ]
		
 KD: Dave.. who would you like to review your published items?

 LH: I intend to do that
		
 DH: LH just sent a mail about status of Handbook..
		
 - Extra item: QAH editing status [2]

 LH: Item (1) The question of how to deal with knock-on effects of TestGL delay.
 there are 6 places in QAH affected.
 I propose that we don't try to rewrite these to reflect delayed status of 
 TestGl but rather put a caveat in status section.
 In ref. section - I would like to explain status of TestGL & point to incubator 
 area/wiki
		
 LH: all links in text point to ref section - which point to TR
		
 KD: but it will be redundant...

 -- [@@ AT I missed some of this conversation]

 LH: Is everyone happy with proposal? [to not rewrite]
		
 [no objections]
		
 LH: Also it lets us talk about it at the next f2f since next publication will be 
 after that.
		
 LH: we agreed to split out appendix - primer & usage scenarios as a seperate doc.
 Follow links on the mail [2] I just sent: item (2). QAH now references QA roadmap. 
 It's unclear what status of roadmap will be.
		
 Regarding items 3,4,5: people might want to look at these and respond by email
		
 LR: Re. QAPD->TSPD: I recal from f2f that we discovered that all different WGs 
 called it a 'test process' document
		
 DH: Using 'TS' makes it too specialized - no room for non-test subjects to be added.
		
 LH: 90% of material is related to tests. but we don't restrict addition of  
 non-test material in there. So maybe we shouldn't clamp down the scope yet.
		
 LH: Re. capitalized question (item 5)?
 DH: 'Yes', should be.
		
 LH: item 4, chronology diagram. this gets more & more irrelevant to QAH. Should 
 we keep it?  (in any form)
		
 DH: I prefer to keep it in some form

 LH: A suggestion to QAWG - can anyone propose a simplfied version?
		
		
 Adjourned


 [1] http://lists.w3.org/Archives/Public/www-qa-wg/2004Jul/0059.html
 [2] http://lists.w3.org/Archives/Public/www-qa-wg/2004Jul/0061.html

Received on Tuesday, 27 July 2004 09:13:18 UTC