That code is still there as you left it. The only thing that has been
deactivated is the creation of range replies from a linear cache miss
(cache hits are still processed in the same way).
From what I can tell you never parsed multipart/byterange in Squid.
You may have done so in test programs, but it does not look like you
ever did this parsing in Squid. At least no such code was ever
committed into the CVS repository.
And yes, I stand corrected. HTTP knows about multipart MIME messages,
but except for the 206 reply it complely ignores them. In 206 replies
proxies MAY decode the multipart content if they like to but the
general message protocol including proxying of 206 replies does not
require understanding of multipart content types and their precense
does not change HTTP operations.
Regards
Henrik
On Wednesday 20 February 2002 23:20, Alex Rousskov wrote:
> When I wrote Range handling code for Squid, the code parsed or
> created multipart responses. I do not know whether that code is
> still in the source tree, but I remember bugging HTTP folks about
> misleading wording in the RFC when it comes to part boundaries.
>
> The primary point of my correction though was to say that HTTP does
> know about multiparts and proxies may want to know as well.
>
> Alex.
Received on Wed Feb 20 2002 - 17:18:53 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:14:48 MST