"Content type" and "media type"

Arms, Caroline caar at loc.gov
Thu Jul 17 17:05:17 CEST 2014


All,



Here is the message, I promised to send to the list on today's call.



After the earlier call where John reported on his action item asking the folks who designed and documented OPC about their use of "content type" and "media type,"  I was motivated to look at RFC 2616 and Part 2 in a bit more detail, in part to try and understand Murata-san's underlying concern.



The explanation John reported is consistent with how I have interpreted the phrase in Part 2 (OPC).  John indicated that the OPC designers saw  "content type" as an abstract concept.  To me that makes sense, since there are lots of typologies for "content" used in different contexts.  In OPC, "media-type" as defined in the RFC is the particular syntactic representation used to identify types of content.   I see a value in having two terms, one for the concept and the other for the particular syntax and instances thereof.  I do not believe that Part 2 gets the usage of the terms quite straight and do think that the definition of "content type" needs adjustment, particularly after the WD0 adjustment in line with ISO guidelines.  [Aside: I don't think RFC 2616 gets its term usage quite consistent either.]



Comparing OPC and RFC 2616:



Each has a data structure with a name based on the words "content" and "type." Both structures incorporate values that follow the media-type syntax.



OPC has the Content Types stream  (defined in 9.2.3.3 in Rex's base WD0 doc) which makes use of a ContentType attribute which takes values that follow the media-type syntax.



RFC 2616 defines a Content-Type entity-header field by:

   Content-Type   = "Content-Type" ":" media-type



=== My personal views ===

1.  We need the two terms but need to review Part 2 carefully for appropriate usage of the terms.  For example, "A content type as defined in RFC 2616." in annotations in the tables in 9.2.3.3.3 and 9.2.3.3.4 seems wrong, since we do not use the Content-Type structure defined in RFC 2616 (see above).



2.  Personally, I have always found the use of "media" in media-type a perplexing use of the word.  I would never use the word "media" to replace "content."  They are not synonyms.



3.  We should avoid major changes in terminology because several formats based on OPC have documentation that makes substantial use of "content type" as a phrase.  Indeed, ECMA 388 (XPS spec.) does not appear to use "media-type" or "media type," only "content type."



Other examples from documentation for OPC-based formats:

"The part's content type property uses a MIME-style media type to describe part content."

"Content Type Component: XML markup (stored in a ZIP item) that identifies the MIME Media Type of each Part in the Package."



4.  The fact that OPC allows "parameters" in the media-type syntax to be disallowed (and XPS does disallow them) seems like another reason to keep the terms distinct.

= = =



If others agree in general, I am prepared (at some convenient future point) to do a careful read of Part 2 with this particular issue in mind and propose changes.



    Caroline


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.vse.cz/pipermail/sc34wg4/attachments/20140717/3eb61d8e/attachment-0001.html>


More information about the sc34wg4 mailing list