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

profiles/modules/levels -- 2 of 2

From: Lofton Henderson <lofton@rockynet.com>
Date: Tue, 22 Apr 2003 13:10:33 -0600
Message-Id: <5.1.0.14.2.20030422130737.03e70050@rockynet.com>
To: www-qa-wg@w3.org


Summary of my opinions on profile/module/level:

1.) combine the 3 concepts?  I'm opposed.

2.) keep separate concepts, DoV and CPs but combine under one GL?  Maybe.

3.) there are good examples of each concept -- archetypes, if you will -- 
and there are examples of misuse.

4.) "4. Definitions" -- we might be able to improve some, but will *never* 
succeed in compact definitions that are sharp enough to completely prevent 
misuse of concepts.

5.) However, with a combination of "4. Definitions", and the discussion and 
examples collected into new sec 2.3, we can start to achieve convergence on 
the use of the concepts.

6.) too much of previous discussion was cut from GL4, 5, 6, and there is 
too little in current (LC) version to disambiguate the concepts.

7.) there are classical or archetypical examples, and we should cite these.

Details...

 From Draft Minutes of 14-April-2003 Teleconference...
At 09:46 AM 4/16/03 -0400, Mark Skall wrote:
>PROFILE/MODULE/LEVEL: (30, 41, 49, 50, 51, 97, 98)
>Some comments suggest that there is not a clear distinction between 
>profile/module/level and that these are all ways to define and label a set 
>of technical requirements.  Can guidelines 4, 5 and 6 be combined?
>
>LR: Should we combine 4, 5 and 6? Sentiment of e-mail was to leave them 
>separate, not to combine them.
>DH: Three concepts adds complexity.  Need distinction among them.   Need a 
>clear definition.
>MS: We need definitions of all of them. However, combining them will make 
>it more confusing

There are three different concepts.  We have good examples of all of them 
in W3C technologies (okay, a little weak on "levels", per my previous mail, 
and we have to go outside of W3C for the best example).  I am opposed to 
trying to combine the concepts.

With the addition of a good section 2.3 (in the new "Concepts" chapter), I 
think that the three GL themselves could potentially be combined into one 
GL.  In combining under one GL, we would preserve the 7 individual 
Checkpoints (4+2+1) and their specificity to the profiles DoV, or to the 
modules DoV, or to the levels DoV.  Note.  It might be possible to combine 
CP4.3, CP5.2, CP7.1 -- the individual CPs about "define relationship of 
this DoV to other DoV -- if we do a good job with the new section 2.Y, 
which will address and discuss the DoV-interrelationship topic.

>DH: Profiles and levels are the same thing.
>LH: Disagree.  They are different but profiles can be levels.  [Ed 
>note.  this is backwards]
>DM: Modules is the only DOV that can be stretched.  It is the generic 
>one.  The other DOVs are well-defined.
>MS: Agree, but how does that help?
>DH: Then we should use different term for "modules"  perhaps "subset."
>[...]
>MS: We need to first define all three, then give reasons for inclusion.
>[...]
>LH: The definitions already exist.
>DH: The definitions are not very good.

Here are the chapter 4 definitions of each term, and two links for each 
term.  I think that the definitions make a reasonable start at defining the 
essential criteria -- i.e., the unique attributes -- of each concept.  They 
could probably be tuned up a bit.

About the links... One link is to the GL section in the current (LC) 
version.  The other is to the GL section in an older version.  I think the 
current version has too little explanation.  The older version probably had 
too much, but did contain additional discussion, comparison, and examples 
that would help provide clarification.

IMO, we are not going to be able to completely separate the concepts with 
"4. Definitions" alone, at least not in such a way that they can't be 
misunderstood and misused.  In fact, I think (heresy!) that we won't 
succeed in getting a definitions that are sharp enough to be foolproof (but 
... this is no different that lots of things that we're dealing with now in 
SpecGL)  However, with Definitions, section 2.3 discussion, and prototype 
examples of the correct application of the concepts, we should be able to 
start creating some convergence in W3C.

module
-----
"4. Definitions" -- a collection of semantically-related elements, 
attributes, and attribute values that represents a unit of functionality. 
Modules are non-hierarchical, discrete divisions that are defined in 
coherent sets.

new: 
http://www.w3.org/TR/2003/WD-qaframe-spec-20030210/#Gd-group-requirements-modules
old:  http://www.w3.org/TR/2002/WD-qaframe-spec-20020826/#b2b3b3d325

Possible archetypes:  XHTML modularization, SMIL, SVG1.1, CSS3, DOM3

(Actually, I have some technical-detail problems with SVG as a good 
archetype.  Not so much with the modularization, as with how the Mobile 
Profiles definitions used it.)

profile
-----
"4. Definitions" -- a subset of a technology that is tailored to meet 
specific functional requirements of a particular application community. A 
profile may address a single technology; or, a profile can also group a set 
of technologies (i.e., from different specifications) and define how they 
operate together. Profiles may be based on hardware considerations 
associated with target product classes, or they may be driven by other 
functional requirements of their target communities.

new: 
http://www.w3.org/TR/2003/WD-qaframe-spec-20030210/#Gd-group-requirements-profiles
old:  http://www.w3.org/TR/2002/WD-qaframe-spec-20020826/#b2b3b3d239

Possible archetypes:  WebCGM, SMIL, XHTML, ...

functional level
-----
"4. Definitions" -- a technology subset that is one of a hierarchy of 
nested subsets, ranging from minimal or core functionality to full or 
complete functionally.

new: 
http://www.w3.org/TR/2003/WD-qaframe-spec-20030210/#Gd-group-requirements-levels
old:  http://www.w3.org/TR/2002/WD-qaframe-spec-20020826/#b2b3b3d443

Possible archetypes:  CGM:1999, various WAI, SpecGL, DOM, CSS

Note.  Per my last email, "DOM Level X", and "CSS Level X" are 
problematic.  Because of the peculiarities of W3C document succession 
(compared, e.g., to ISO CGM), they don't qualify as DoV under our LC-66 
issue resolution.

All for now,
-Lofton.
Received on Tuesday, 22 April 2003 15:16:22 GMT

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