Summary:
Single brackets are getting some troublesome false positives in Facebook's
install. Particularly, there's a weird convention at Facebook of tagging diffs
by putting stuff like "[perf]" or "[chat]" in the title, although this isn't
turned into structured data at any stage. When commits appear in Diffusion, we
currently link such ad-hoc tags to Phriction.
Wikipedia uses double-bracket sytnax, as do many other wikis, so this seems like
a reasonable burden to place on the lightweightness of the markup. The
alternative is selectively disabling Phriction markup in some interfaces, but
I'd rather allow integration in commit messages and just guard the syntax more
closely.
(I'm not providing any sort of migration plan since this landed less than a week
ago and I'm pretty confident no one has built a huge wiki yet, but I added a
CHANGELOG note.)
Test Plan: Edited a wiki document and added some links. Verified single brackets
were unlinked and double brackets were linked.
Reviewed By: jungejason
Reviewers: hsb, aran, jungejason, tuomaspelkonen
CC: aran, jungejason, epriestley
Differential Revision: 689
Summary:
Major things taking place here:
- A new table for storing <task, project> relationships.
- Moved all task query logic into a dedicated class.
- Added a "projects" filter to the UI.
I was originally going to try to drive this off the main search index but the
perf benefits of a custom schema make an overwhelming argument in favor of doing
it this way.
Test Plan:
Filtered tasks by author and owner and zero, one, and more than one project.
Exercised all the group/sort options. Ran the index script over my 100k task
corpus. Edited task-project membership and verified the index updated.
Reviewed By: cadamo
Reviewers: gc3, jungejason, cadamo, tuomaspelkonen, aran
CC: aran, cadamo, epriestley
Differential Revision: 556
Summary:
Conduit already has multiple-session code, just move it to the main
establishSession() method and set a web session limit larger than 1.
NOTE: This will log everyone out since we no longer look for the "web" session,
only for "web-1", "web-2", ..., etc. Presumably this doesn't matter.
Test Plan:
Applied patch, was logged out. Logged in in Safari. Verified I was issued
"web-1". Logged in in Firefox. Verified I was issued "web-2".
Kept logging in and out until I got issued "web-5", then did it again and was
issued "web-1" with a new key.
Ran conduit methods and verified they work and correctly cycled session keys.
Reviewed By: tuomaspelkonen
Reviewers: tuomaspelkonen, jungejason, aran
Commenters: jungejason
CC: rm, fzamore, ola, aran, epriestley, jungejason, tuomaspelkonen
Differential Revision: 264
Summary:
When I tested this earlier I was incorrectly interpreting PHPMailer errors as
SES errors. This works fine as long as you get around the peculiarities of
PHPMailer.
Test Plan:
Sent email to myself, received email from a human-readable address in my mail
client.
Reviewed By: rm
Reviewers: rm, tuomaspelkonen, jungejason, aran
CC: aran, rm, epriestley
Differential Revision: 246
