Re: <XML:SCRIPT>

W. Eliot Kimber (eliot@isogen.com)
Sat, 11 Jul 1998 10:29:03 -0500

At 07:14 AM 7/9/98, Peter Murray-Rust wrote:
>At 21:57 08/07/98 -0500, W. Eliot Kimber wrote:
>[...]
>>If somebody wanted to write the two-page spec for the mechanism, I'd be
>>happy to learn Java sufficient to write a reference implementation.
>
>Agreed. My impression is that the nitty-gritty of associating FPIs/URLs,
>etc with code is:
> - not terrifyingly difficult (like many others I have built SGML catalog
>software)
> - unlikely to come directly out of W3C efforts
>in which case it could be an excellent thing for XML-DEV to run with. The
>important thing is to agree communally on how to do it. I would see this as
>a logica next step after XSchema.
> I hope my mail didn't seem negative - I'm delighted to use NOTATION IFF:
> - other people are keen on using it
> - we agree on a software spec
> - there is a communal implementation

Cool--I didn't mean to be testy in my initial responses--just in a bad mood
generally last week.

The think about notations as opposed to MIME types is that notations are a
more general mechanism, not tied to any particular use domain, while MIME
types are designed specifically to make things work smoothly on the
Internet. Nothing wrong with that, but I just prefer the more general
solution as a rule (but I'm like that). Because MIME types can be
associated with notations (say by using the MIME type as the public or
system identifier for the notation), the two mechanisms can be complimentary.

Ideally, XML and SGML would have a mechanism for declaring specialized
forms of notations, which would better reflect the way we use notations in
the HyTime standard, for example. In HyTime, we defined several distinct
types of notations: queries, storage managers, etc. It would be nice to be
able to do something like:

<!-- Declare a notation type: -->

<!DCNTYPE -- Data content notation type declaration --
QUERY
PUBLIC "this is the pointer to the authoritative definition of this
type of notation"
>

<!-- Now use the notation type to declare a specific notation: -->
<!QUERY SDQL PUBLIC "ISO/IEC 10179//NOTATION Standard Document Query
Language//EN"
>

The "<!QUERY" declaration is still just a notation declaration, but it is
more clearly specialized.

I don't know if the value of this type of specialization is worth the cost,
but it seems like a reasonable thing to do, in the abscense of any deep
thought about the issue.

Cheers,

E.

--
<Address HyTime=bibloc>
W. Eliot Kimber, Senior Consulting SGML Engineer
ISOGEN International Corp.
2200 N. Lamar St., Suite 230, Dallas, TX 75202.  214.953.0004
www.isogen.com
</Address>