OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: PSVI in XML



At 11:53 AM 4/16/01 +0800, Rick Jelliffe wrote:
>There is no XML representation that I know of that augments the document.
>(I would expect Extensibility have one, because of their Datatypes for DTDs
>note at http://www.w3.org/TR but it is not worked out for the current spec.)

While DT4DTD does define an Information Set Contribution 
(http://www.w3.org/TR/dt4dtd#AEN92), it doesn't go so far as defining what 
it would look like in XML.  I'd be surprised if XML Schema used the 
approach defined in DT4DTD a-dtype declarations, but that could be 
interesting in its own way.

>Martin mentioned Richard Tobin's dump format.  It is a great thing, but
>unsuitable for use over the web because of blowout.

I think that's what I'd heard of, and I'll investigate it further.  It 
doesn't sound precisely like what I'm looking for, but a start is better 
than none.

>To my knowledge, there is no effort from any group to define an XML
>representation of the PSVI that maintains the original document's structure.
>So every distributed process which needs to use the infoset has to re-create
>the PSVI infoset from the raw document and its schema.

Yep.  Feels like an enormous amount of repeated work to me.  I know there 
are some people who claim that the infoset is all, but there aren't a lot 
of good ways to distribute infosets over networks in easily processed 
forms.  Apart from XML, anyway.

>It is interesting to ponder how this gap impacts the claim that XML Schemas
>will allow us to move to a distributed architecture of piped-and-typed
>trees.

Certainly!  For me, it's more about finding alternates routes to reach the 
PSVI, and for processing the PSVI after initial PSVI creation.  The current 
validation story seems to me to do too much all at once - validation, type 
labeling, defaulting, etc. - but doesn't leave a lot of room for doing 
other things with that information.

It'll also be interesting to see how the DOM folks deal with all of this.


Simon St.Laurent - Associate Editor, O'Reilly and Associates
XML Elements of Style / XML: A Primer, 2nd Ed.
XHTML: Migrating Toward XML
http://www.simonstl.com - XML essays and books