Leo Franchi 49328c2551 Unify rabbitmq usage to use only one queue. ... To work around the issue we're having with queue draining between parallel blocking connections, use the same rabbitmq queue for both activity and presence events, keyed on a 'type' flag in the message itself. (imported from commit 188e8fda1695734e52c5740db2195072cfc81479)		2013-02-15 11:43:17 -05:00
..
bugdown	Don't match ** for strong across newlines	2013-02-14 11:56:53 -05:00
__init__.py	Add simple parallel computation library.	2012-09-14 13:32:25 -04:00
actions.py	Unify rabbitmq usage to use only one queue.	2013-02-15 11:43:17 -05:00
avatar.py	Factor out Gravatar hash calculation	2012-10-17 01:09:16 -04:00
bulk_create.py	Move action functions from models.py to zephyr/lib/actions.py.	2013-01-11 16:11:03 -05:00
cache.py	tornado: Cache messages internally in tornado process.	2013-01-11 16:11:07 -05:00
context_managers.py	Manage file locking using the 'with' statement	2012-11-16 13:43:44 -05:00
initial_password.py	Generate initial API keys in the same manner as passwords	2012-10-17 18:23:01 -04:00
message_cache.py	tornado: Fill the message memcached in a child process.	2013-01-14 12:28:51 -05:00
parallel.py	Remove unused imports	2012-10-25 15:22:18 -04:00
query.py	Add a Django query helper	2012-09-28 19:33:15 -04:00
queue.py	Handle SIGINT and SIGTERM in process_user_activity by closing communication gracefully	2013-02-15 11:05:31 -05:00
response.py	Move json response functions into their own file	2012-11-08 16:30:57 -05:00
timeout.py	timeout: Try to kill the thread once the timeout expires	2013-01-30 11:13:48 -05:00
timestamp.py	Rename zephyr.lib.time to zephyr.lib.timestamp.	2013-01-11 15:58:19 -05:00