[ moving to squid-dev ]
From what I can see, the site is using JavaScript to do autocomplete
on a search field. The autocomplete requests use POST, but without a
body.
With Firefox, this results in a POST request without a body; i.e., it
doesn't have transfer-encoding *or*
7441122 wrote:
squid-3.1.0.8-20090613
Fedora 11
./configure -with netfilter ( only)
make all
/usr/src/squid-3.1.0.8-20090613/src/CommCalls.cc:13: undefined
reference to `cbdataInternalLock(void const*)'
/usr/src/squid-3.1.0.8-20090613/src/CommCalls.cc:13: undefined
reference to
This patch adds some polish and minor unfinished business in IPv6
support vs delay pools.
Also updates the bucket size to 64-bytes to allow for 4MB traffic caps.
Still need to bump up the refresh size for 4MB/sec flows.
Amos
=== modified file 'src/DelaySpec.cc'
--- src/DelaySpec.cc
Bundle Buggy has detected this merge request.
For details, see:
http://bundlebuggy.aaronbentley.com/project/squid/request/%3C4A331D05.5060601%40treenetnz.com%3E
Project: Squid
Hi,
Here are some tiny patch to fix cross-compilation of squid 3.0.13. Here
host is x86_64-pc-linux-gnu and dest is i586-geode-linux-uclibc.
There is a first issue with cf_gen. cf_gen is generated with TARGETCXX
and should be build with HOSTCXX as it is used on runtime compilation.
Weird. The code that I'm assuming generates the 411 is (squid2-HEAD):
if (!clientCheckContentLength(request) || httpHeaderHas(request-
header, HDR_TRANSFER_ENCODING)) {
err = errorCon(ERR_INVALID_REQ, HTTP_LENGTH_REQUIRED, request);
http-al.http.code = err-http_status;
On Thu, 18 Jun 2009 10:20:09 +1000, Mark Nottingham m...@yahoo-inc.com
wrote:
Weird. The code that I'm assuming generates the 411 is (squid2-HEAD):
Reporter is using 3.0. The header parse and validation code is similar only
regarding flow and design.
Particular lines and if-else are very
