Re: ACTION-892: Information Conformance Statement for CT Guidelines

On Tue, 23 Jun 2009 14:32:00 +0200, Rotan Hanrahan  
<rotan.hanrahan@mobileaware.com> wrote:

> Just commenting from the sidelines... I find the version with SHOULD and  
> MUST together to be the better version. While I can understand that the  
> SHOULD levels present the implementers with "grey areas" to consider,  
> and therefore a greater demand for guidance/explanation, seeing these  
> clauses excerpted in this manner does not provide sufficient context for  
> a proper understanding.

Agreed - for my 2c.

> The tabular format is rather good, I must say. However, I cannot see the  
> reason why one would permit a comments cell for the MUST clauses. In  
> such cases, failure to comply with a MUST means total non-compliance,  
> and I would not offer a "comments box" for excuses. Perhaps those cells  
> should be greyed out, and the Introduction amended so as to stress that  
> comments only apply to explanations for non-compliance with the SHOULD  
> or SHOULD NOT clauses.

I don't think it matters if they are there - in essence you just need a  
single checkbox for MUST, and 'yes' or 'why not' for should - although  
being able to note a failure in a seperate column (with explanation of it)  
is helpful. But I don't think you need to spend a lot of time on that.  
Implementors can edit the table if they really need to for internal use...

cheers

Chaal

> My 2c worth.
>
> ---Rotan
>
> PS The text of 4.1.5.4 looks odd.
>
>
>
> -----Original Message-----
> From: public-bpwg-request@w3.org [mailto:public-bpwg-request@w3.org] On  
> Behalf Of Francois Daoust
> Sent: 23 June 2009 12:55
> To: Mobile Web Best Practices Working Group WG
> Subject: ACTION-892: Information Conformance Statement for CT Guidelines
>
> Hi,
>
> Per my ACTION-892, please find below links to two different versions of
> the Implementation Conformance Statement that should ship with the
> guidelines.
>
> The action was:
> [[
> Prepare an ICS with MUST/MUST NOT (to view if that's a good idea), try
> to add a "depends on" column, explain "Not applicable" or remove it.
> ]]
>
> I chose to remove the "Not applicable" column. I agree it merely created
> confusion with no useful outcome.
> I haven't tried to add a "depend on" column. I am not sure how to do it,
> and wonder whether that would be of any use in our case anyway.
>
> The ICS is generated automatically from the spec (through an XSLT
> stylesheet). The excerpts are atomic, i.e. one line per normative
> statement, but that means some sentences needed to be cut into pieces,
> and some excerpts do look meaningless without context.
>
> I prepared two versions of the ICS that matches the latest draft of the
> guidelines: one that contains both SHOULD-level and MUST-level
> statements (this includes SHOULD NOT, RECOMMENDED, and the rest of the
> tribe, of course), and another one that only contains SHOULD-level
> statements. The rationale for the ICS was to have exceptions to the
> SHOULD statements explained, and so we had initially restricted the
> statements to put in the ICS to SHOULD-level statements only. I must say
> that I now find the "full" version with MUST-level statements as well
> more useful.
>
> Version with SHOULD-level and MUST-level statements:
> http://www.w3.org/2005/MWI/BPWG/Group/TaskForces/CT/editors-drafts/Guidelines/ics-090622-must
>
> Version with SHOULD-level statements only:
> http://www.w3.org/2005/MWI/BPWG/Group/TaskForces/CT/editors-drafts/Guidelines/ics-090622
>
> (Please ignore the Abstract and Introduction sections for the time
> being, they do need to be rewritten, in particular for the version that
> contains SHOULD-level and MUST-level statements)
>
> For dizcussion/rezolution:
> - "SHOULD and MUST" or "SHOULD only"?
> - publish the ICS directly within the guidelines or as a separate  
> document?
> - comments to improve the table?
>
> Francois.
>


-- 
Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

Received on Thursday, 25 June 2009 11:52:54 UTC