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

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

From: Larry Masinter <lmm@acm.org>
Date: Wed, 1 Dec 1999 08:22:14 PST
To: "Chris Lilley" <chris@w3.org>, "MURATA Makoto" <murata.makoto@fujixerox.co.jp>
Cc: "Dan Connolly" <connolly@w3.org>, <timbl@w3.org>, <simonstl@simonstl.com>, <ietf-xml-mime@imc.org>, <Tsmith@parc.xerox.com>, <xsl-editors@w3.org>
Chris Lilley wrote:
> > > a) application/xml for xml files. All are required to be well 
> formed,as
> > > per the XML specification, otherwise it is a fatal error.
> > > b) application/xml-epe for external parsed entities which are not
> > > themselves well-formed instances

MURATA Makoto replied:

> > Point of clarification.   I have an external parsed entity.  I use
> > it only as an external parsed entity.  Incidentally, it also parses
> > as an XML document, but this was never intended.  Which one should
> > I use?

Chris Lilley wrote:

> The algorithm is as follows:
> Is it well formed instance
>   Yes: use a)
>   No: use b)

This is not an appropriate algorithm. The choice of the MIME type
to use for a communication depends on the intent of the communication
as well as the nature of the body of the communication. It might
be appropriate to use text/plain, for example, if you intend for
the recipient to read the actual XML rather than interpret it,
and it might be appropriate to use application/octet-stream if you
want no interpretation at all.

The point is that the general question of "this message body, which
MIME type should I use" can always have multiple answers (since
application/octet-stream usually applies, for example).

The XHTML group got caught up in this, since there are some XHTML
bodies for which 'text/html' are appropriate and others that are not.
Don't try to construct a functional one-to-one mapping from 
message body content to MIME type, because you can and will get
multiple values.
Received on Wednesday, 1 December 1999 11:21:34 UTC

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