W3C home > Mailing lists > Public > www-math@w3.org > May 2000

Re: MathML in http://www.imc.org/draft-murata-xml

From: Murata Makoto <mura034@attglobal.net>
Date: Mon, 08 May 2000 23:54:17 +0900
Message-Id: <200005081454.AA02523@makoto.attglobal.net>
To: David Carlisle <davidc@nag.co.uk>, www-math@w3.org
Cc: simonstl@simonstl.com, Dan Kohn <dan@dankohn.com>, mura034@attglobal.net

Thank you for your mail.  We certainly appreciate your comments. 

> I note in the XML media draft at the URL above, the section:
> 7.13 application/mathml-xml

We are willing to update the example or even omit it if you like.  

There is one important issue.  Do you really need a specialied media type?  
If users typically embed MathML expressions in XML documents, the MathML 
expressions are not MIME entities and thus do not have associated media 
types.  Dispatching programms have to rely on namespaces.  If MathML 
expressions are typically embedded, it might not be a good idea to introduce 
a specilized media type.

It MathML expressions often become standalone MIME entities, it makes 
sense to introduce a specialized media type.  By doing so, you can avoid 
fetching irrelevant XML, you can introduce MathML-specialized fragment 
identifiers, and you can register your MathML browsers to MIME engines.

If you really need a specialized media type and you have special 
concerns (such as specialized fragment identifiers), it might 
be a good idea to remove this subsection and create another I-D 
dedicated to MathML.  In particular, if you need specialized 
fragment identifiers, it is not a good idea to follow the naming 
convention suggested in our I-D.  If not, the naming convention 
is recommended.  It allow the use of generic XML processing to 
MathML data.

> As you may know, MathML2 is currently in `Last Call' so the time for
> making any changes/additions is rapidly running out.
> Could you confirm what the current status of this proposal is, I note
> the draft says
>    This Internet-Draft will expire on May 31, 2000.
> which isn't long:-) so I assume another draft is on the way.

Yes, we are preparing a significantly modified version.  We will 
publish another I-D very soon (in a few days, I hope), but we are 
expecting at least another I-D.

> If MathML is about to have an official MIME type sanctioned, it ought be
> mentioned in the mathml spec....
> Actually I'd have thought that text/* would be more suitable, certainly
> if using the file with a mathml unaware application I'd rather see the
> MathML source markup, which is the distinction between application/* and
> text/* given in the introduction to your document.

In the IETF-XML-MIME ML, there have been a lot of discussion about 
when to use text/*.  The top-level media type "text" has been frequently 
misued, but it is intended to be readable for casual users.   Thus, 
our latest draft has a recommendation as below:

   If an XML document -- that is, the unprocessed, source XML document
   -- is readable by casual users, text/xml is preferable to
   application/xml. MIME user agents (and web user agents) that do not
   have explicit support for text/xml will treat it as text/plain, for
   example, by displaying the XML entity as plain text. Application/xml
   is preferable when the XML MIME entity is unreadable by casual
   users. Similarly, text/xml-external-parsed-entity is preferable when
   an external parsed entity is readable by casual users, but
   application/xml-external-parsed-entity is preferable when a plain
   text display is inappropriate.

Ned Freed (co-author of MIME RFCs) wrote very useful mail about this 
issue.  It is available at:


> For your information, the current MathML2 draft (chapter 7) says
> <p>MIME types offer an alternative strategy that can also be used in
> current user agents to invoke a MathML renderer.  This is primarily
> useful when referencing separate files containing MathML markup from
> an <kw role="element">EMBED</kw> or <kw role="element">OBJECT</kw>
> element.  The W3C Math Working Group suggests that generic MathML be
> assigned the MIME type <code>text/x-mathml</code>, and for browser

x-* is probably not a good idea.  RFC 2048 is quite clear about that 
(see below).

   For convenience and symmetry with this registration scheme, media
   type names with "x." as the first facet may be used for the same
   purposes for which names starting in "x-" are normally used.  These
   types are unregistered, experimental, and should be used only with
   the active agreement of the parties exchanging them.

   However, with the simplified registration procedures described above
   for vendor and personal trees, it should rarely, if ever, be
   necessary to use unregistered experimental types, and as such use of
   both "x-" and "x." forms is discouraged.

> registry, we suggest the standard file extension <code>.mml</code> be
> used.  In MathML 1.0, <code>text/mathml</code> was given as the
> suggested MIME type.  However, the assignment of MIME types to XML
> applications has come into question in the interim.  Thus, beginning
> with MathML 2.0, we suggest instead using the less-regulated
> experimental MIME type <code>text/x-mathml</code>.</p>

Is text/mathml registered at IANA?

> This message is BCC'ed to the internal Math WG list, I am not sure if
> you both have W3C access, if so you may reply to that list, otherwise
> I'll forward any replies back to the group. Alternatively you could use
> the public mailing list being used for last call comments on mathml
> www-math@w3.org

As a member of the XML Core WG, I have access.  But my co-authors do not. 
I am ccing to the public comment ML.


Internet: mura034@attglobal.net
Nifty: VEQ00625
Received on Monday, 8 May 2000 10:56:25 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 20 February 2010 06:12:49 GMT