> Jeff, do you have an RFC reference for the format and
type mimetype keywords?
> I could swear I've seen them but can't find it for the
life of me.
I don’t think there was anything as specific as a
definition for format and type. There are these general statements on Content-Type
parameters from RFC 2046:
After the type and subtype names, the remainder of the header field is simply a set of parameters, specified in an attribute/value notation. The ordering of parameters is not significant.
Parameters are modifiers of the media subtype, and as such do not fundamentally affect the nature of the content. The set of meaningful parameters depends on the media type and subtype. Most parameters are associated with a single specific subtype. However, a given top-level media type may define parameters which are applicable to any subtype of that type. Parameters may be required by their defining media type or subtype or they may be optional. MIME implementations must also ignore any parameters whose names they do not recognize.
I think this leaves the definitions for format and type up
to us (the DITA TC). I suggest that they should be defined to be the same as or
a sub-set of the values that are allowed for the attributes of the same name on
the topicref element.
Back in 2006 I said that all of the parameters should be
optional, but I am now wondering if format and charset sould be required and
type should be optional.
Do we want an optional navtitle parameter?
Does application/dita+xml have to mean documents that use
DITA specialization or could it mean documents that conform to the DITA
specification? The later would allow us to use application/dita+xml with
format=ditaval and eliminate the need for a separate MINE type registration for
From: Erik Hennum
Sent: Friday, April 18, 2008 8:57
Subject: [dita] mime type for DITA
DITA-Minded Technical Committee:
I thought it might be useful to refresh the thread about mime types with a
summary of the proposal prior to the meeting.
This proposal is really a light modification on Jeff's analysis. (Jeff, do you
have an RFC reference for the format and type mimetype keywords? I could swear
I've seen them but can't find it for the life of me.)
Please note the DISCUSSION item below about the type.
= A mime type for DITA
Where client tools needs to download documents from a CMS (potentially via
WebDAV) and route requests to open those documents to the desktop editor that
understands DITA specialization.
Such client tools needs to distingish DITA topics and maps (with base or
specialized vocabularies) from other XML documents (XMI, SVG, what have you)
provided by the CMS system. The CMS typically stores the document without an
extension, so the tool can't rely on file-system mechanisms for recognizing
DITA vocabularies. If the tool simply passes through the document from the CMS
to the client Operating System, the web brower might open the XML topic instead
of the DITA-aware editor. Without a MIME type, the tool will have to open and
parse the initial part of the file to recognize DITA topics and maps, which
will be very inefficient.
The mime type record (such as an HTTP header) has the following fields:
* The application/dita+xml mime type identifies any document that use DITA
* The format property identifies the base vocabulary for the specialized
document (in particular, topic or map).
* The type property identifies the specialized vocabulary.
The character set and encoding fields have no DITA-specific considerations.
NOTE: Topics of ditabase don't have a separate format because they are just
topic documents with a list of topics rather than a single topic.
NOTE: Because it doesn't have specialization, the DITA values file requires a
separate mime type of application/ditaval+xml. If it ever becomes specialized,
the mime type of application/dita+xml and format of ditaval would be
DISCUSSION: The type property could reflect either:
* The type of the root topic, first topic (for ditabase), or map by using the
class attribute (replacing spaces with a legal separator character)
* The name of the shell
The case for using the topic type: Fallback processing becomes possible for the
root type prior to opening the document. If the document is declared as
javaClass and the receiver understands apiClassifier, the receiver knows that
it can process the vocabulary before opening the document. If the receiver
cares about domains or subtopics, however, the receiver must open the topic to
The case for declaring the shell: If the receiver knows the shell, the receiver
understands everything about the document without opening it including the root
type, nested topics or topics after the first for ditabase as well as domains
in either topic or map. If the receiver doesn't know the shell identifier,
however, the receiver would have to open the document to determine fallback