proposal on Review Assignments

QAWG --

Please comment on this proposal for Review 
Assignments.  Yes?  No?  Modify?  Please comment also on proposed sign up 
date (7/27).

 From the minutes of today's telecon...

At 11:56 AM 7/10/02 -0400, Karl wrote:
>[...]
>6.) Review assignments [2]
>         [2] http://www.w3.org/QA/Group/2002/06/reviews [Group Only]
>
>LH We will do a case study wrt to the QA Framework
>... each person will do 3 reviews in the next 6 months
>... sandra does it seem reasonnable (you did XML)
>
>SM it takes 2 hours in the morning
>
>LH it seems reasonnable
>... you look at the guidelines and checkpoints and tries to extract the 
>techniques
>
>Karl is asking for clarification
>
>LH After the teleconf, I will write a clear proposal and detailed.
>... we can discuss it on the list.
>... and we can start to put names in a Matrix for assignments
>-----

Okay, here is my clarification, and proposal.

Background:  see [1], which is some discussion between Mark and I.  I 
(probably) misunderstood Mark's proposal at Montreal.  In [1] I explain why 
I don't think it is optimal, and outline the "3+1" modified proposal.

Proposal:
====

Part 1:
---
Each QAWG member will do three case studies over the next 4-6 months -- one 
for OpsGL, plus one for SpecGL, plus one for TestGL.  You'll start with a 
skeletonized version of the GL document that looks something like 
[2].  You'll fill in comments about how the target (WG, spec, TS) relates 
to the checkpoint.  (Note.  There will be a bit of summary front matter for 
you to provide also, and a bit of per-checkpoint front matter -- the final 
skeleton that you use will contain placeholders.)

Those of us who have done them for OpsGL so far think that they will take 
around 1/2 day each, on the average, if you pick a WG/spec/TS that you're 
familiar with.

The rationale is two-fold:

1-1.) this is critical for us, QAWG, in order to develop good GL 
documents.  We who did OpsGL found that some bits of OpsGL didn't work so 
well when you tried to actually apply it.  Also, each QAWG member will have 
a really thoughtful look at each GL part, which seems valuable to us for 
quality GL development.

1-2.) it generates ample raw material for the Extech parts.

Part 2:
---
Each QAWG member will do one "techniques analysis" over the next 4-6 months 
-- either one for OpsGL, or one for SpecGL, or one for TestGL.

Rationale:  This will be the material from which we build the Extech 
"Techniques" bits (which will still be complemented and supplemented by 
some Case Studies [Examples] materials).

Details:

2-1.) we need to moderate these "techniques analysis" assignments, to make 
sure that we get even coverage of Ops, Spec, Test.  I.e., we want 2-3 
members to cover each one, instead of 7 Ops, 2 Spec, 0 Test.

2-2.) Timing.  We could use Ops now, e.g., before mid-August, so that we 
could have a OpsExtech document cycle before Tokyo (and in any case, before 
anticipated October publication of Frm parts).  Spec would be good before, 
say, before Tokyo.  Test [...uncertain...]

2-3.) Labor.  Unknown.  We haven't done it yet.  Likely more than a case 
study.  I would expect the Lead Editor of the Extech part to be one of the 
2-3 volunteers.  In fact, he/she could subdivide the ckeckpoints of the GL 
part amongst the 2-3 volunteers, to reduce per-member labor requirement.

Signup:
=====

Assuming QAWG approval of something like this, would it be reasonable to 
request WG members to sign up by next telecon (7/27)?  Signup is handled by 
sending a message to Dom, per [3].  For the case studies (Ex), I don't see 
much constraint on what you choose (although duplication dilutes rationale 
1-2 above).  For techniques analysis (Tech), we may shuffle people around a 
bit to even out the coverage.

Regards,
-Lofton.

[1] http://lists.w3.org/Archives/Public/www-qa-wg/2002Jun/0065.html
[2] http://www.w3.org/QA/WG/2002/06/ops-skeleton-sample.html
[3] http://lists.w3.org/Archives/Public/www-qa-wg/2002Jun/0031.html

Received on Wednesday, 10 July 2002 19:07:23 UTC