Re: Proposed changes to address issue-dbooth-3 (ambiguity)

>
> Booth, David (HP Software - Boston) wrote:
>   
>>   
>>     
>>> From: public-grddl-wg-request@w3.org 
>>> [mailto:public-grddl-wg-request@w3.org] On Behalf Of Clark, John
>>>
>>> [ . . . ] with the proposed change, the
>>> transformation author can only control the variability along 
>>> one of two
>>> possible axes; the other is out of her control.  First, the document
>>> must be projected into the XPath data model in order to determine what
>>> transformations apply (call this the "transformation determination"
>>> step).  Second, the document must be consumed by each of the GRDDL
>>> transformations and the results merged (call this process the
>>> "transformation application" step).  The implementation of these two
>>> steps is independent, and the transformation author can only 
>>> exert full
>>> control over the transformation application step (for example by using
>>> XProc, as we have theorized).
>>>     
>>>       
>> Yes, that is exactly correct.  The proposed change does not permit all
>> variability to be controlled, it only permits the "transformation
>> application" step to be controlled.  In this regard, this proposal would
>> be clearly inferior to a change that would also remove the variability
>> of that step.
>>     
It seems to make the most sense that since a GRDDL agent is a
client-side process, that if one is concerned about whether a
GRDDL-aware agent has variability in the first step (i.e. finding the
transformation in the source document), then one can simply use a
GRDDL-aware agent that permits the local user to control that
variability. However, it seems unneeded to have to tell the
implementation authors that they should let their end-users control such
variability, except perhaps as informative advice.

However, the problem that the GRDDL transformation cannot be unambiguous
(although, it may in fact be ambigous if the transformation language
permits this ambiguity) without  giving it the whole representation (the
*second step*) is a distinct issue completely separate from the issue of
controlling variability in the GRDDL-aware's agent parsing of the source
document.
>> Hmm, with that in mind, perhaps I should offer at least one other
>> alternative solution.  I'll send it in a separate message.
>>
>>
>> David Booth, Ph.D.
>> HP Software
>> +1 617 629 8881 office  |  dbooth@hp.com
>> http://www.hp.com/go/software
>>
>> Opinions expressed herein are those of the author and do not represent
>> the official views of HP unless explicitly stated otherwise.
>>  
>>
>>
>>   
>>     
>
>
>   


-- 
		-harry

Harry Halpin,  University of Edinburgh 
http://www.ibiblio.org/hhalpin 6B522426

Received on Tuesday, 19 June 2007 15:53:26 UTC