Factoring out Content-Disposition (i123), was: Content-Disposition (new issue?)

Hi,

we discussed this issue during the WG meeting in Dublin 
(<http://jabber.ietf.org/logs/httpbis/2008-07-29.txt>), and it seems 
there was some support for the idea of taking Content-Disposition out of 
HTTP/1.1.

My own proposal was to clearly specify how Content-Disposition works in 
existing HTTP clients, and to move that into a separate, 
standards-track, spec.

However, an essential part of supporting Content-Disposition is to 
understand a subset of RFC 2231, which defines  extensions to MIME 
header parameters (<http://greenbytes.de/tech/webdav/rfc2231.html>), for 
instance specifying how to transport non-ASCII characters in C-D's 
filename parameter. *That* mechanism is independent of C-D, and could be 
usable for other specs, such as the link header spec (which uses 
parameter notation as well).

So we actually need *two* documents:

1) A clarification of what parts of RFC 2231 apply to HTTP/1.1 
(essentially a profile), and

2) A specification of Content-Disposition in HTTP, built on top of RFC 
2183 and the profile above.

I've got a proposal ready for the first document -- it's available at 
<http://greenbytes.de/tech/webdav/draft-reschke-rfc2231-in-http-latest.html>. 
The plan is to submit a -00 draft soon, but I'll be more than happy to 
incorporate feedback that arrives before I do.

BR, Julian

Received on Thursday, 14 August 2008 21:02:19 UTC