Re: new draft xml proc profiles doc

On Thu, Dec 15, 2011 at 1:16 PM, Henry S. Thompson <ht@inf.ed.ac.uk> wrote:
> James Fuller <jim@webcomposite.com> writes:
>
>> Admittedly they are more for human parsing, but as they precede the
>> profile's precise definition I thought this was the right place for
>> rationale to go as any clarification can be quickly addressed.
>
> To me they appear so terse as to be perplexing, rather than helpful,
> to the casual reader.
>

I based the selection of text on 'least objection' minimums, as we
will have different reactions depending on who consumes this document.
I decided it was best to support existing titles, as they experienced
a lot of debate in the past, which was about as far as I was prepared
to go.

> I had in mind to just put them at the beginning of section 2, just
> before section 2.1, perhaps in the following form:

as previous email, I think we reduce usefulness by moving the text
away from the profiles formal description. Perhaps there is some
benefit in obfuscating in one place and making a simple overview area
in another, but I'm not overly convinced. Practically I think it justs
forces readers to bunny hop around the document.

>  The four profiles defined here identify four increasingly rich
>  profiles, in terms of kinds of processing and amount of information
>  provided to applications:

I would put this in section 2.0 ;)

>  The Basic profile, adding only support for xml:base processing to
>  the minimum required by the XML specification, in order to allow for
>  correct resolution of relative URIs;
>
>  The Id profile, which adds xml:id processing in order to identify
>  IDs in the possible absence of complete attribute type declaration
>  information;
>
>  The External Declarations profile, which adds mandatory external
>  markup declaration processing in order to guarantee all
>  information-affecting declarations are processed;
>
>  The Full profile, which adds xi:include processing, in order to
>  transclude linked infosets as parsed XML or as text, recursively as
>  required.

I like how you massaged the descriptions …. shall I flow them in now
and we can debate the merits of location at todays telcon ?

J

Received on Thursday, 15 December 2011 12:37:55 UTC