W3C home > Mailing lists > Public > www-qa-wg@w3.org > January 2003

Re: My action items AI-20030108-2 and AI-20030108-3 (Test Task Force Charter) completed

From: Lofton Henderson <lofton@rockynet.com>
Date: Fri, 17 Jan 2003 10:49:24 -0700
Message-Id: <5.1.0.14.2.20030117094751.02bd6ec0@rockynet.com>
To: Dimitris Dimitriadis <dimitris@ontologicon.com>, www-qa-wg@w3.org

Dimitris,

Thanks for getting this out well ahead of schedule.  I have a couple of 
comments on your concerns about "Mission", as well as the on draft itself.

At 05:46 AM 1/17/03 +0100, Dimitris Dimitriadis wrote:
>[...]
>All minuted changes from the F2F were incorporated (except for the part on 
>templates of test materials, which I cannot remember having either voted 
>for or abstained from voting on during the F2F according to an earlier 
>email to the list).

See my earlier message about this.  (In fact, I don't think we actually 
voted on anything during this discussion.)


>For the archives, I again want to stress that I do not agree that we 
>should strive for not enforcing that test techniques as defined and 
>discussed in other QA WG materials be used.

I don't think we resolved to NOT enforce (or seek enforcement).  We 
resolved to state the mission as "ensure", and IMO enforcement is one of 
many tactics or techniques that can be used to ensure that the practices 
are followed.  We can and should use both the carrot and the 
stick.  "Enforce" is nearly synonymous with "stick" (and there would be a 
number of specific sticks, i.e., enforcement techniques that we could use).

>Our current wording (on which we reached consensus during the F2F) is 
>"ensure". I cannot, on the one hand, understand what makes "ensure" 
>different from "enforce",

See above comments.  "Ensure that guidelines are followed" is a general 
statement of our goal.  "Enforce" is one way to "ensure".

>except if "ensure" means that we just check for something having been done 
>by someone else (which means that in order to ensure, someone needs to 
>have enforced, except for cases where all gets done voloutarily); on the 
>other I think this is merely words and that it does not really make that 
>big a difference.

I disagree.  Part of the negative reaction to "enforce" is that it is too 
much "in your face".  It immediately defines us as the QA Cops.  Another 
problem with "enforce" is that it is too limited and specific.  There will 
likely be enforcement, but as I said before, this is only a part of the 
picture.  "Inducement" is another part.


>Again, I updated the draft as per the minutes, but I want it archived that 
>I do not agree with this weaker wording. Some group needs to enforce that 
>things be done, if not this group then some other. However, as this is my 
>personal view and not that of the QA WG, the TTF charter does not reflect that.

I think you are misinterpreting the Seattle sentiment (at least, 
misinterpreting why I myself wanted the word "enforce" changed).

Here's another way to look at it.  We have resolved (see Issues List) that 
the Framework GL documents should become mandatory.  This presumably 
includes TestGL.  Doesn't "mandatory" imply "enforce"?

As LR pointed out also -- see Seattle minutes -- we in QA don't actually 
have enforcement power, and most likely never will.  We can promote a 
manditoriness rule, and maybe even embed it in our GL documents like I18N 
did.  But then it will be the role of the Director or Comm or WebMaster or 
Chairs or ... to enforce the rule.

Specific comments on draft:

Mission statement, editorial.  Replace "QA Testing guidelines Document" 
with the correct title "'QA Framework: Test Guidelines' specification".  Or 
else make it a generic and less limiting reference, "QA test guidelines 
documents".  (I guess I favor the latter).

Duration.  This is unchanged from the previous draft (except for the 
addition of the last sentence):

>Depending on whether the TTF is to be a separate body from the QA WG 
>within the current QA activity, or if it is to coexist within the activity 
>but as a separate group, either of two durations are relevant:
>
>     * The duration (including the possibility of rechartering) of the 
> current QA WG
>* A duration which reflects the need for a TTF group in the W3C, meaning 
>that it be chartered with the general W3C needs in mind (for example, 
>two-year charter periods, renewable if necessary)
>
>The TTF is anticipated to be active for at least two years, regardless of 
>form.

By definition, the TTF is a part of the QAWG.  We have discussed and 
resolved (Tokyo and Seattle both), that we will not attempt to start a 
separate TWG at this point.  I.e., the most expedient way to actually start 
doing useful test work is to form the TTF, start work, recruit help.  Then 
at re-charter we either flip-flop so that main mission and deliverables of 
QAWG are test (and Framework is a task force), or spin off a new TWG, or 
whatever (depends a lot on how many people we can attract to "test", what 
other stuff happens in QA and in W3C in the next 8 months, etc).

Accordingly, I (hastily) propose this for Duration:

"The TTF will be formed and operate initially within the QAWG, with both 
existing staffing and additional staff to be solicited.  It is anticipated 
that the mission of the TTF will require at least two years.  At the time 
of QAWG rechartering (August 2003), a number of options will be 
considered.  One possibility is that the mission of a re-chartered QAWG is 
redefined to give emphasis and priority to the TTF mission and 
deliverables.  Another possibility is that the TTF is spun off as a 
separate Working Group."

Thoughts (anyone)?

-Lofton.
Received on Friday, 17 January 2003 12:47:03 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:13:12 GMT