This patch enables logging of all call attempts from a queue. Not only the ones that do not fail.
The patch also introduces a new parameter "congestion" to both RINGNOANSWER in queue_log and AgentRingNoAnswer AMI event, which is set to 1 if the call failed to go through because of technical difficulties.
This makes it easier to make queue_log statistics with information about problems with an agent. For example if an agent has a faulty line, or your telco/dahdi connection is having problems.
I am however unsure if everyone want this marked as an congestion from the "AST_CONTROL_CONGESTION" frame. Since in my experience, this can come if a SIP UA doesn't want to let you ring more than x seconds, etc. Most real congestion problems come before this frame is generated. (read: before a new channel is up at all) So if this patch should be applied, maybe it should be configurable, or let out.