W3C home > Mailing lists > Public > xsl-editors@w3.org > October to December 1999

Re: External parsed entities (Re: Inconsistency between IETF and W3C...)

From: MURATA Makoto <murata.makoto@fujixerox.co.jp>
Date: Fri, 03 Dec 1999 20:29:45 +0900
Message-Id: <199912031129.AA03494@archlute.fujixerox.co.jp>
To: ietf-xml-mime@imc.org
Cc: Dan Connolly <connolly@w3.org>, timbl@w3.org, simonstl@simonstl.com, Tsmith@parc.xerox.com, xsl-editors@w3.org
Chris Lilley wrote:

> Ok, so even allowing that, if "which program" is an XML parser and you
> feed it a non well formed document, what happens? I would regard
> "mandatory fatal error" as being synonymous with "can't read the
> document"

Right.  But even if the content type is text/xml-external-parsed-entity, 
I cannot read it.  I have no programs that can handle external parsed 
entities directly.  But such programs may be developed tomorrow.

> Yes. Even more, though, I care about the converse case; in the case of
> applications/xml, it should always be the case that it is xml. If it is
> not, then it is an error, rather than being "a little used case which is
> actually legal but we dodn't expect people to use it directly".

We cannot guarantee that the MIME body is really XML since errors will happen.  
But we can elminate the possibility that correctly-labelled data cause 
fatal errors.

> > If some programs that can handle external parsed entities directly (i.e.,
> > without using XML processors for XML documents that reference to them),
> > your point perfectly makes sense. 
> Your point seems to ignore resource discovery, indexing, and other
> routes to finding resources that may not have been intended to be found,
> or found first, or found before other documents in whose context they
> were intended to be used.

I do not ignore such discorvery.  My point is that even if it is lablled as 
an external parsed entity, no existing programs can parse it anyway.  But 
this may change.

Having heard your argument, I am inclined to introduce a media type 
for external parsed entities.  How do other people feel?

As for external PARAMETER entities, they always parse as external DTD subsets.  
I now think that we do not need a specialized media type for external parameter 


Fuji Xerox Information Systems
Tel: +81-44-812-7230   Fax: +81-44-812-7231
E-mail: murata.makoto@fujixerox.co.jp
Received on Friday, 3 December 1999 06:26:59 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:44:17 UTC