Re: TestGL Issues list (alpha release)

At 06:06 PM 2/5/04 -0800, Patrick Curran wrote:
>Attached is a preliminary issues list for TestGL.

Excellent!  I haven't read the details, but from the issue count alone, it 
looks like it must be a pretty comprehensive start.

>It incorporates all of the comments from minutes of our discussions in 
>Boulder (even when I couldn't understand what these meant) plus additional 
>feedback from Dimitris, Mark, and myself.

The mention of Boulder (Oct 2003) made me curious...

Question:  how far back do the collected issues go?

What I'm thinking about is the extensive dialog about TestGL that was 
started by J.Carroll (early summer), and kicked back and forth between him 
and Karl.  He, for example, was the ultimate source of the discussion about 
"waterfall".  (You'll find the email dialog in www-qa and/or www-qa-wg 
archives).

My concern is this:

1.) we need to ensure that we observe proper protocol for acknowledging 
comments and/or replying to their substance (to ensure that we are not 
vulnerable to attack for process violations);

2.) therefore, the issue list should take care to acknowledge an 
originator, esp. if he/she is outside of the QAWG.

Now having said that, I'm not sure what our formal Process obligations are, 
to JC about TestGL issues raised -- the dialog I mentioned was about an 
early, pre-LC Working Draft.  But ... even if we don't consider him to be 
the direct or most recent source of an issue -- i.e., perhaps we (QAWG) 
synthesized the issue from loose www-qa discussion -- we should have some 
attribution in the Description if he was involved in the topic.

I do *not* recommend that we go back and mine those old comments for every 
identifiable detailed issue, especially since TestGL has evolved 
substantially since then (indeed, in part because of those comments).  I 
think that would be over-kill.  (Other views?)

JC is the only extra-QAWG case like this that I can think of, 
offhand.  This is not critical right now, but let's put it on the work 
queue for TestGL issues list, for next-version refinement.

>I modelled the format of the table on our SpecGL issues list, but tweaked 
>it slightly (hopefully you'll see the value of the changes I've made).

Is the HTML your editing source?  Or are the editable issues data in XML, 
which you transform to HTML?

-Lofton.


>Please review the issues, and send feedback. I'm particularly interested 
>in clarifications of the questions embedded within the text (most of these 
>are preceded '[PC]') but you should also feel free to respond to the 
>comments from [MS] and [DD].
>
>Of course, additional issues will also be welcomed.