Commentate and reorganize ansible.cfg, adding one or two quality-of-life options along the hway

ansible.cfg

@@ -1,18 +1,49 @@
 [defaults]
+# I have a large number of machines, which warrants a large forks setting
+# here.
 forks = 16
+# We set gathering to smart here as I'm often executing the site-wide playbook,
+# which means a ton of redundant time gathering facts that haven't changed
+# otherwise.
+gathering = smart
+# host_key_checking is disabled because nearly 90% of my Ansible plays are in
+# ephemeral environments and I'm constantly spinning machines up and down.
+# In theory this is an attack vector that I need to work on a solution for.
 host_key_checking = false
+# Explicitly set the python3 interpreter for legacy hosts.
 interpreter_python = python3
+# Default inventory is labeled "production", but this entire environment is a
+# homelab so whatever.
 inventory = inventories/production
+# I hate fun
+nocows = true
+# Pipelining is a fantastic speedup
+pipelining = true
 private_key_file = ~/.ssh/desu
+# Set explicitly here, but this should be the default.
 remote_tmp = $HOME/.ansible/tmp
 remote_user = ansible
+# This extra roles directory is used for Galaxy roles.
 roles_path = .roles:roles
 system_warnings = true
 
+[diff]
+always = true
+
+[inventory]
+# I have no clue why this isn't the default
+unparsed_is_failed = true
+
 [privilege_escalation]
+# We force an always-become condition because of the myriad of things we need
+# to do as root.
 become = true
 
 [ssh_connection]
-pipelining = true
+# The number of retries here is insane because of the volatility of my home
+# network, where a number of my machines live.
 retries = 15
+# These extra args are used for bastioning, where the ephemeral Ansible
+# controller remotes into a bastion machine to access the rest of the
+# environment.
 ssh_extra_args = -o ForwardAgent=yes -o StrictHostKeyChecking=no