Dear media-types-expert, Dear Björn,

Forwarded-from : member-math@w3.org From : "Robert Miner" <robertm@dessci.com> Date : 24 novembre 2009 03:43:50 GMT+01:00 To : <member-math@w3.org> Archived-At: <http://www.w3.org/mid/D1EFB337111B674B8F1BE155B01C6DD6034CCA68@franklin.corp...

...

Dear Björn,

I am writing in response to your comments on the media-type registrations for MathML given in Appendix B of the Last Call draft of the MathML 3 Specification. Your original comments are recorded at http://www.alvestrand.no/pipermail/ietf-types/2009-October/ 002268.html.

We have modified Appendix B, and believe we have addressed the issues you noted. The changes are explained below, and you can see the draft changes in an editor's draft of the spec located at

http://monet.nag.co.uk/~dpc/draft-spec/appendixb.html

Since we need to record the resolution of all Last Call comments, could we ask you to let us know whether you accept these changes as addressing your comments? Thanks in advance.

The specific actions taken are:

...

The first problem I have with this is that there is no word on what exactly the various types are for, like if there are any restrictions on what should be in a mathml-content+xml vs a mathml-presentation +xml document, or if the types can be used with MathML 2.0.

A new introductory section entitled "Selection of Media Types for MathML Instances" was added to specifically address this issue. The related text in section 6.2.3 of the spec is essentially summary in nature, and was therefore left unchanged. It already referred to appendix B for detailed information on the usage of the media types, and that information is now given.

The section carefully defines the presentation and content MathML vocabularies, and gives a clear rule for determining what media type should be used for a given MathML instance based on the contents of that instance.

The section also contains some explanation and examples to motivate the usage of the three MathML media types. The section ends by addressing use of the media types with earlier versions of MathML

...

As per RFC 3023, the charset definition and encoding considerations should be referenced as follows, which the proposals fails to do:

Registrations for new XML-based media types under top-level types other than "text" SHOULD, in specifying the charset parameter and encoding considerations, define them as: "Same as [charset parameter / encoding considerations] of application/xml as specified in RFC 3023."

These sections have been edited to match the proper format.

...

The security considerations are missing (in fact the specification as a whole does not have a security considerations section either).

Security considerations have been added. We have listed both the generic issues common to similar XML languages, and two issues specific to MathML: the possible presence of executable code in semantic annotations, and the (possibly inadvertant) denial of service risks arising in computational contexts from trying to solve unsolvable problems.

...

I believe the text you have under "interoperability considerations" is misplaced there in all three cases.

This text has been completely rewritten. Again, there are some generic considerations common to similar XML languages, as well as several MathML specific issues. In particular, lack of versioning information in MathML instances introduces a backward compatibility concern, and the result of evaluating MathML expressions is not guaranteed to be the same in different computational systems.

...

I note that under "Applications that use this media type" you have "(todo)". Going to Last Call with "todo" markers left is not a good practise. I note that the purpose of this field is to give a general idea of what kind of applications use it, not to list individual software products.

Thank you for the clarification on the purpose. We have filled out the section appropriately.

...

Under "Person & email address to contact for further information" the proposal fails to properly separate name and email address, use something like "Name <address>" instead. I do not think having a generic W3C / W3C Webmaster combination there is a good practise

Following other media type registrations under the BCP13 process, we have changed this to list Paul as the person, with the Math WG group mailing list for the address, followed by a pointer to the public W3C Math web site for additional information.

--Robert