* social:
Better watching for story changes. This is both a speed improvement and usability fix, since now stories that haven't changed recently but did change before will no longer be declared as updated every time they are fetched. Props to @textfiles for the push.
* social:
Auto-linkifying comments and replies, and stripping html from comments, replies, and story titles, tags, and authors.
Adding compacting to user sub folders, to remove dupe subscriptions.
Clearing folders and decoding unicode.
Typo in gunicorn conf.
Only using appropriate number of gunicorn processes, making this machine agnostic.
Deferring OPML import if it takes > 20 seconds. Email user when complete. Also, hamburgers. This one's for @torrez.
Adding vps roles to fabfile.
Conflicts:
apps/reader/views.py
apps/social/models.py
* social:
fixing bug with feed exception when there is a duplicate feed id
Fixing shift+a for social subs. Also fixing bug where site would stay selected after adding a new site.
Fixing tutorial by forcing dialog resize. Thanks to @afita for finding this bug.
Adding necessary cookie domain settings to local_settings template.
Whoops, forgot to remove 1 second sleep from autocomplete.
Fixing the living snot out of the add site dialog. Adding an autocomplete spinner, showing results on focus, and no longer auto-adding on click (allowing user to select a folder). Big thanks to @afita for the push.
Fixing broken offset for locked mouse indicator on site reload. Thanks to @afita for bug report.
Removing cjson. Fixing collect_stats requirement on bootstrap.
Adding logging to celerybeat. Adding site unread count calculation to homepage freshening.
Moving cronjob tasks over to celerybeat. Starting with task_feeds, collect_feedback, collect_stats, and mark_read for homepage freshening.
* master:
Being much more aggressive about certain sites trying to take over the iframe. Blocking nytimes.com, twitter.com, and stackoverflow.com from even allowing an Original view.
Conflicts:
media/css/reader.css
* master:
Fixing icon fetching error when the page is corrupted.
Writing a glorious hack that forces a count of unread scores.
Conflicts:
apps/reader/models.py
apps/rss_feeds/management/commands/refresh_feeds.py
fabfile.py
utils/feed_fetcher.py
* master:
Changing the color of sql duration times in debug logging. Yeah, nobody but me will see this, but boy is it pretty. Ask me sometime and I'd love to show you my insanely colorful logs.
Correctly handling story that lack titles but have content. Using a parsed and shucked truncation of content in place of missing titles.
Sanity check on deleted feeds.
Read it Later -> Pocket
Read it Later -> Pocket
Conflicts:
media/js/newsblur/reader/reader.js
* master:
Showing feed pushes in statistics.
Instrumentation to count feed pushes.
Migration to fix starred stories with incorrectly double-encoded url strings.
Fixing bug around missing feed links being checked for length.
Conflicts:
apps/rss_feeds/views.py
* master:
Adding MongoDB logging for debug.
Being far kinder during feed exception handling and showing users error text. Cannot believe this wasn't there.
Conflicts:
media/css/reader.css
templates/base.html
utils/json_functions.py
