Re: Introduction - pre patch submission

2010-05-03 Thread Amos Jeffries
On Mon, 03 May 2010 15:24:30 -0430, Robert Marcano wrote: > On 05/03/2010 03:07 PM, Henrik Nordström wrote: >> mån 2010-05-03 klockan 21:59 +1200 skrev Amos Jeffries: >> >> Donẗ forget to add whildcard interface name syntax such as ppp+ or >> similar, as used for many dialup or vpn network configu

Re: Introduction - pre patch submission

2010-05-03 Thread Henrik Nordström
mån 2010-05-03 klockan 15:24 -0430 skrev Robert Marcano: > I understand, but we are using tcp_outgoing_address because it supports > ACLs and we route some users to a dedicated ISP exclusive to them. Perfectly valid reason. > It breaks server side persistent connections so we disable it, but I

Re: Introduction - pre patch submission

2010-05-03 Thread Robert Marcano
On 05/03/2010 03:07 PM, Henrik Nordström wrote: mån 2010-05-03 klockan 21:59 +1200 skrev Amos Jeffries: Donẗ forget to add whildcard interface name syntax such as ppp+ or similar, as used for many dialup or vpn network configurations sharing the exact same dynamic ip problem. But generally my r

Re: Authentication based on ip address

2010-05-03 Thread Henrik Nordström
mån 2010-05-03 klockan 15:58 +0100 skrev Michael Graham: > My current feeling is that I will need to send the IP address to the > helpers. This would mean putting the IP address into the > AuthUserRequest and then adding a configuration option to the send the > address. My preference would be to

Re: Introduction - pre patch submission

2010-05-03 Thread Henrik Nordström
mån 2010-05-03 klockan 21:59 +1200 skrev Amos Jeffries: > So an automatically generated outgoing IP address based on an iface name > then? > Selecting the iface, then picking from one of its addresses? > > Sounds like it should actually work. Donẗ forget to add whildcard interface name syntax

Re: [MERGE] Allow testheaders.sh to keep testing after a failure

2010-05-03 Thread Henrik Nordström
sön 2010-05-02 klockan 20:32 +0200 skrev Kinkie: > Ideally testheaders.sh should be rewritten to be done using make targets. > Would allow for parallelization. > Problem is, it's hard to do it portably. GNU make has lots of handy > extensions :( So does automake, for removing the need of fancy ma

Authentication based on ip address

2010-05-03 Thread Michael Graham
Hi all, I'm interested in being able to authenticate to different LDAP servers or to only allow transparent authentication from some IP addresses. I've had a brief look at the wiki[1] but it doesn't seem to work correctly with NTLM authentication. My current feeling is that I will need to send t

Re: Introduction - pre patch submission

2010-05-03 Thread Amos Jeffries
Robert Marcano wrote: On 04/30/2010 10:42 PM, Amos Jeffries wrote: Robert Marcano wrote: - Make tcp_outgoing_address be able to use an interface name and not only a fixed ip address, this solve a problem we have with some setups where we allow to acces to the internet with a dedicated ISP for

Re: [PATCH] Accept ICAP OPTIONS responses with unknown body types.

2010-05-03 Thread Amos Jeffries
Alex Rousskov wrote: Accept ICAP OPTIONS responses with unknown body types. Warn about the unknown OPTIONS body type but ignore it instead of rejecting the entire OPTIONS response. Note that ICAP does not standardize OPTIONS body types, and Squid does not recognize any body type. This patch is

Re: [MERGE] Allow testheaders.sh to keep testing after a failure

2010-05-03 Thread Kinkie
>> Ideally testheaders.sh should be rewritten to be done using make targets. >> Would allow for parallelization. >> Problem is, it's hard to do it portably. GNU make has lots of handy >> extensions :( >> > > Aye.  +1 on the commit anyway as an immediate temporary measure. Trying a resend, ezmlm re

Re: [MERGE] Allow testheaders.sh to keep testing after a failure

2010-05-03 Thread Amos Jeffries
Kinkie wrote: On Sun, May 2, 2010 at 8:12 PM, Alex Rousskov wrote: Do not stop testheaders.sh on the first error, to be compatible with "make -k check". If testheaders.sh stops on the first error, it becomes very difficult to test your changes with "make check" when somebody else broke some ot