* master: (48 commits)
Updating copy on homepage.
Marking all as read on Everything view should clear folder unread counts.
Fixing error on committing to cache for rate limits when the dummy cache backend is used.
Adding linux.com story to press page.
Uniqifying stories when read.
Moving from sha module to hashlib.
Syncing time to account for time drift on VPS.
Fixing issue around showing too many Oh Nos on abort.
Save hash for address+link migration. necessary for a unique index.
Deprecating db02, moving redis and rabbitmq over to db01.
Updating nginx conf to support better amintenance pages giving a 503 status code.
Moving NewsBlur counts to beginning of title.
Fixing that crummy missing existing story issue.
Updating feedparser.
Fixing broken hashlib issue with feeds and no link.
Adding MongoDB slave to local_settings template.
Adding two new columns, necessary for the feed_settings branch: hash of address+link, and locking address.
Marking read stories in Feed view as very much read. Gradientizing, too.
Fixing /m/ for no good reason.
File under "fucking finally": Adding client-side error handling and callbacks for feeds and river.
...
Conflicts:
config/nginx.newsblur.conf
* master:
Fixing broken signup form for google reader import.
Adding /load_feeds to repeat visitor tracking, since the iPhone app never hits refresh_feeds.
Adding celery granularity to fabfile.
* master:
Changing color of logged user agent.
Adding browser user agent to all logging.
Fixing sizing on iphone app's original view's url.
Only active feeds get loaded in the iPhone app. It's about time.
Showing correct story_id in logging for bad read stories.
Showing HUD on logout in iphone app.
Completing the python api.
Showing the correct logging info for mismarked read stories.
Fixing river of new to use new user stories.
Re-running story read migration to account fornew read stories during the migration process.
Adding errors to API/login and API/signup. Getting rid of namespacing on login and signup form fields for api use.
Moving read stories from story_date based to story_id based.
Handle error when pkilling pgbouncer during pgbouncer bouncing.
Adding in Python API. Needs some love to get the rest of the api up to standard.
