QA Working Group Teleconference Monday, 07-June-2004 -- Scribe: Karl Dubost Attendees: (dd) Dimitris Dimitriadis (Ontologicon) (KD) Karl Dubost (W3C, WG co-chair) (JR) Patrick Curran (Sun) (DH) Dominique Haza�l-Massieux (W3C - Webmaster) (LH) Lofton Henderson (CGMO - WG co-chair) (LR) Lynne Rosenthal (NIST - IG co-chair) (MS) Mark Skall (NIST) (DM) David Marston (IBM, QA IG) Regrets: Absent: (AT) Andrew Thackrah (Open Group) Summary of New Action Items: Karl: to send issues lists for SpecGL by Wednesday Lofton: to send issues lists for QAH by Wednesday Agenda: Previous Telcon Minutes: Minutes: Form to apply to the QA WG Dom: Please be sure to fill the form for the QA WG http://lists.w3.org/Archives/Public/www-qa-wg/2004Jun/0000.html http://www.w3.org/2004/01/pp-impl/34282/join QA WG F2F Draft Agenda Lofton: the agenda is quite sketchy now. I would like that editors of SpecGL and TestGL come with a draft agenda. Karl: I will draft a few issues list for SpecGL list: AI. Dimitris: Yes, I will do that for TestGL AI. I would love to have 9am TestGL topic if I'm not able to participate physically. Lofton: I will send an email to invite reviewers to discuss the new versions of the documents (SPecGL). Lynne: and for QAH Lofton: I'll do Strategy QA WG: Lofton: We will have to discuss the strategy for the QA WG. So please be ready to discuss about that at the meeting. TestGL http://lists.w3.org/Archives/Public/www-qa-wg/2004Jun/0014.html http://lists.w3.org/Archives/Public/www-qa-wg/2004Jun/att-0014/TestGL-20040607.html Lofton: Dimitris, you will present TestGL MS: Could you hightlight what has specifically changed between the two versions. Dimitris: I will do that. (presenting Conformance testing) Do we have a definition of conformance testing. MS: yes in the glossary: Karl? Karl: Yes, please dimitris review it. http://www.w3.org/QA/glossary Dimitris: (Dimitris is explaining a few changes) * mapped back to the Specs * test repeatable and reproduceable MS: are they in the QA glossary? Karl: No. Lofton: I would support that it's one place or the other, spec or glossary and we can migrate them after. Dimitris: in one central place Lynne: But for now on I would love to have it in the back of test gl so to be able to have it in one document. Dimitris: * Mandatory items: test, metadata, test reviews, author, etc. * It is meaningful to put a requirement for reproduceability and reproducatbility (2.5) over time? We have tried to follow-up on the comments of Jeremy Caroll MS: Dimitris: Lynne: Patrick has sent an outline of what should be in a test or what should be in a test suite. we should point to a few things like that. Dimitris: We have designed issues which look like normative. Dom: We don't need to make the technique normative. Lynne: You just stay away from the requirements and you made them Good Practices or techniques. Dimitris: Ok I can do. At least, I can make them good practices. we have to rewrite in something more generic. Test Design issues. I have introduced some wording about Bug tracking report. DM: Do you address how to deal with changes made in the errata document of a specification. Dimitris: I'm more on the side, this is a bunch of tests, run them and tell us what you have as results. 3.3 A test suite has to be maintained after the life of the WG. So how to deal with that. * this is basically the changes that have been made to the documents. I have tried to separate process and what is belonging to the tests Dom: we could keep them into the document by making normative and non normative. Dimitris: ok. What about people saying we are duplicating. Dom: if we can link to where it comes from. Patrick: We should have something for test Author Dimitris: If we put together people will say that we should not tell them how to do it. Lynne: it's not necessary mandatory, but it will be helpful for other people. Karl: You can create a starter kit where you say to people you can follow these steps to create a test suite. Dimitris: I will try to introduce under the design part of the document, I will introduce Best Practices. * Submissions must be reviewed. Information in the test metadata. * Issues must be tracked in a formal manner [@@ Has been adressed above] * Test materials must be formally tested (by you, by your users) Lofton: What will we do on the document from now on? Lynne: We need work on this document before to be published. Karl: It needs also a bit of improvement on the layout. Lofton: There's a need of Technical Work Lynne: Patrick, Dimitris, Dig one section and don't worry about others. karl: Just work deep on one section, a small one, and we will be able to figure out what will be the organization. So what's requirements, techniques, examples, and things like that. I will provide the HTML markup for it and the CSS. and it will help to figure out what's normative and not. Lynne: +1 Dimitris: ok. Adjourned