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

Re: AI-2003-1-13-1 - Checkpoint 9.4 rewording - done

From: Lofton Henderson <lofton@rockynet.com>
Date: Wed, 22 Jan 2003 07:32:32 -0700
Message-Id: <5.1.0.14.2.20030122072430.036cb1e0@rockynet.com>
To: Karl Dubost <karl@w3.org>, www-qa-wg@w3.org

Thanks for the clarifications, Karl.

Actually, I have had a change of mind since yesterday.  Rather than revisit 
in telecon, let's leave it to the editors (and any other interested 
parties) to work out the wording.

Reason:  In last week's telecon, the issue itself was closed (AR-25) with 
general agreement that the original "standard way" needed clarification, 
and in telecon we even agreed more or less what we were trying to say.  Now 
we just have to sort out how to say it.

A couple more comments embedded...

At 08:36 AM 1/22/03 -0500, Karl Dubost wrote:

>At 17:05 -0700 2003-01-21, Lofton Henderson wrote:
>>>Checkpoint 9.4. Define a abstract mechanism to create extension [Priority 3]
>>>
>>>To fulfill this checkpoint a specification MUST provide a unique way of 
>>>defining the extension, each time is it authorized by the specification. 
>>>It is not applicable if extensions are not allowed.
>>
>>I'm unclear whether this captures what I thought we meant by "standard 
>>way to define extension", or whether on the other hand it changes the meaning.
>
>what you don't understand I guess is the word abstract, I guess. I have 
>used it with the intended meaning of model. So if it's unclear for you, 
>there's a slight chance it's unclear for others.

Yes, that was the first thing that stopped me.


>for example in CSS3, it would be the fact to have a mechanism which 
>permits extension but only in this way
>         -vendor-propriety: value;
>Where the extension must always start with a dash.
>Where after the dash you must have the vendor's name, for example: moz
>and finally the usual semantics of propriety with the right characters.

I agree.  SVG has similar mechanisms (e.g., it requires that any 
implementation-defined extension elements or attributes be in a different 
namespace).


>This is an abstract mechanism or model as you wish, which is define for 
>all cases. It's theoretical.

In email or with the editors, we can sort through it some more.  I was also 
troubled by "unique" -- in the minutes, we used the word "uniform", but the 
minutes don't capture all of the detail of the discussion (you said it 
really well (IMO) at one point, but now I can't remember the details).

-Lofton.
Received on Wednesday, 22 January 2003 09:30:15 GMT

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