On Jul 30, 2007, at 6:52 AM, Henrik Nordstrom wrote:
On mån, 2007-07-30 at 01:16 -0700, Ricardo Newbery wrote:
As far as I can tell, PURGE requests via HTTP/1.1 works fine. But
I've read some documentation that implies that this was not always
the case. The release notes don't seem say. Anyone know which Squid
version started supporting PURGE via HTTP/1.1 ?
It's always worked I think..
What do not work well is PURGE of Vary objects (objects having a Vary
header).
Regards
Henrik
Okay. FYI, I'm rewriting some documentation (and fixing code) for a
tool that generates PURGE requests so I would like to list any edge
cases where special care might be warranted.
Because of the broken implementation of HTTP/1.0, I would like to
switch the tool to using HTTP/1.1 (which works much better in my
tests). So if I understand correctly, Squid PURGE has always
supported the HTTP/1.1 protocol and this change shouldn't break
anything. Hmm... I wonder why the original authors of the tool
seemed to believe otherwise.
Regarding PURGE of Vary objects. The solution is just not to include
a Vary in the PURGE, correct? So all variants are purged.
Ric
