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

Re: SpecGL editorial comments

From: <david_marston@us.ibm.com>
Date: Fri, 19 Dec 2003 16:06:58 -0500
To: www-qa-wg@w3.org
Message-ID: <OFE2EEDDFD.3A73584B-ON85256E01.0071981A@lotus.com>

LH>> CP3.1, Examples:  I cannot figure out what is meant by "...only
LH>>one profile can be implemented at a time."

DHM> Well, you could imagine that it would be forbidden to implement
DHM> both a tiny and a mobile profile in an SVG implementation (why
DHM> would you do that, I don't know, but maybe that helps understanding
DHM> the meaning :)

Thanks, Dom, that does help. But I think it might make this question
into a substantive one. Perhaps the "example" paragraph could be
expanded to say "profiles are intended to be mutually exclusive"
instead of "only one ... at a time" and that this limitation might
characterize any given instance of the data/consumer/whatever. I
don't believe anyone is trying to say that mutual exclusivity is a
necessary characteristic of all profiles, just an optional one.

Here's an example of how it gets tricky. A content producer may offer
to provide content at any of several profiles. Some of these may not
exceed others in all aspects, like offering higher-res visuals but no
sound. But let's say there is at least one pair of profiles, call them
the "wristwatch" and "PDA" profiles, where we can say that PDA exceeds
wristwatch in all aspects. If the producer asks a PDA whether it can
handle the wristwatch profile, the PDA should say yes. This is true for
the agile producer I postulated but also for a wristwatch-profile-only
producer. The constraint about who *implements* more than one profile
just gets in the way, whereas a constraint that any given instance of
the content must conform to exactly one profile may be mildly useful.

LH>> CP9.3, Discussion:  what does "relevant assuring parties" mean?

DHM> e.g. a certification authority?

I guess so, or an indepedent test lab or a magazine. Actually, I think
it's always "claimants" because the checkpoint is stating that it is
not just the implementer (nor just a certification authority) who can
make a claim. Do the "relevant assuring parties" somehow magnify the
degree of responsibility that binds the claimant?

LH>> Sec 4.5:  notice the nasty formatting glitch of the blockquote in
LH>>the Conformance Claim example (unless your browser window happens
LH>>to be just the right width).

It definitely looks worse under Netscape than Internet Explorer.
(A good software tester keeps more than one browser available.)
.................David Marston
Received on Friday, 19 December 2003 16:07:20 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:13:15 GMT