W3C home > Mailing lists > Public > public-xml-processing-model-comments@w3.org > January 2010

Re: Associating HTTP headers and bodies

From: Florent Georges <fgeorges@fgeorges.org>
Date: Thu, 7 Jan 2010 23:18:35 +0100
Message-ID: <ebaca5bf1001071418t7bfd65f7i9d25dc2f9a84e208@mail.gmail.com>
To: Norman Walsh <ndw@nwalsh.com>, Alex Milowski <alex@milowski.org>
Cc: public-xml-processing-model-comments@w3.org
2010/1/7 Norman Walsh wrote:

  Hi Norm,

>> So before going further, is it a typo in the content model, is
>> it me not understanding the content model or the prose, or is
>> it something else?

> Total screw-up by the editor. That got left behind after we
> cleaned things up. My new understanding is:

> 1. The c:headers in a c:multipart are for the multipart
>    message.

  Weird.  Why not simply use c:request/c:header then?

> There's no way to add any other headers and if there are any
> other headers on a multipart reply, they get dropped on the
> floor.

  So the content model of c:multipart is really (c:body)+,
instead of the current (c:header*,c:body+), right?

  About RFC 2387 only allowing Content-* headers for parts of a
multipart entity, I cannot find this requirement in the RFC.  I
can only find examples with Content-*, but not the rule forbiding
other headers.  I guess it is in another MIME RFC (this one is
only for multipart/related).  Did I miss anything?

  Regards,

-- 
Florent Georges
http://www.fgeorges.org/
Received on Thursday, 7 January 2010 22:19:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 7 January 2010 22:19:09 GMT