Summary:
A few similar requests have come in across several tools and use cases that I
think this does a reasonable job of resolving.
We currently send one email for each update an object receives, but these aren't
always appreciated:
- Asana does post-commit review via Differential, so the "committed" mails are
useless.
- Quora wants to make project category edits to bugs without spamming people
attached to them.
- Some users in general are very sensitive to email volumes, and this gives us
a good way to reduce the volumes without incurring the complexity of
delayed-send-batching.
The technical mechanism is basically:
- Mail may optionally have "mail tags", which indicate content in the mail
(e.g., "maniphest-priority, maniphest-cc, maniphest-comment" for a mail which
contains a priority change, a CC change, and a comment).
- If a mail has tags, remove any recipients who have opted out of all the
tags.
- Some tags can't be opted out of via the UI, so this ensures that important
email is still delivered (e.g., cc + assign + comment is always delivered
because you can't opt out of "assign" or "comment").
Test Plan:
- Disabled all mail tags in the web UI.
- Used test console to send myself mail with an opt-outable tag, it was
immediately dropped.
- Used test console to send myself mail with an opt-outable tag and a custom
tag, it was delivered.
- Made Differential updates affecting CCs with and without comments, got
appropriate delivery.
- Made Maniphest updates affecting project, priority and CCs with and without
comments, got appropriate delivery.
- Verified mail headers in all cases.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran, epriestley, moskov
Maniphest Tasks: T616, T855
Differential Revision: https://secure.phabricator.com/D1635
your own actions
Summary:
- Mail.app on Lion has cumbersome threading rules, see T782. Add an option to
stick "Re: " in front of all threaded mail so it behaves. This is horrible, but
apparently the least-horrible option.
- While I was in there, I added an option for T228.
Test Plan:
- Sent a bunch of threaded and unthreaded mail with varous "Re:" settings,
seemed to get "Re:" in the right places.
- Disabled email about my stuff, created a task with just me, got voided mail,
added a CC, got mail to just the CC.
Reviewers: btrahan, jungejason
Reviewed By: btrahan
CC: aran, mkjones
Maniphest Tasks: T228, T782
Differential Revision: https://secure.phabricator.com/D1448
Summary: There are lots of callsites to $changeset->getFilename() so it seemed
easier to rename getFileName() to getFilename() even if it includes database
change. Plus I think that getFilename() is better.
Test Plan:
Alter database.
Open revision.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran
Differential Revision: https://secure.phabricator.com/D1437
Summary:
- For context, see T547. This is the last (maybe?) in a series of diffs that
moves us off raw sha1() calls in order to make it easier to audit the codebase
for correct use of hash functions.
- This breaks CSRF tokens. Any open forms will generate an error when
submitted, so maybe upgrade off-peak.
- We now generate HMAC mail keys but accept MAC or HMAC. In a few months, we
can remove the MAC version.
- The only remaining callsite is Conduit. We can't use HMAC since Arcanist
would need to know the key. {T550} provides a better solution to this, anyway.
Test Plan:
- Verified CSRF tokens generate properly.
- Manually changed CSRF to an incorrect value and got an error.
- Verified mail generates with a new mail hash.
- Verified Phabricator accepts both old and new mail hashes.
- Verified Phabricator rejects bad mail hashes.
- Checked user log, things look OK.
Reviewers: btrahan, jungejason, benmathews
Reviewed By: btrahan
CC: aran, epriestley, btrahan
Maniphest Tasks: T547
Differential Revision: 1237
Summary: See T625. Facebook's REST-based MTA layer had a check for this so I
overlooked it in porting it out. We should not attempt to deliver email to
disabled users.
Test Plan:
Used MetaMTA console to send email to:
- No users: received "no To" exception.
- A disabled user: received "all To disabled" exception.
- A valid user: received email.
- A valid user and a disabled user: received email to valid user only.
(Note that you can't easily send to disabled users directly since they don't
appear in the typeahead, but you can prefill it and then disable the user by
hitting "Send".)
Reviewers: btrahan, jungejason, nh, tuomaspelkonen, aran
Reviewed By: aran
CC: skrul, aran, epriestley
Differential Revision: 1120
Summary: Allow tweaking Differential mail before sending.
Test Plan:
Wrote a listener renaming Differential attachments and it worked without
problems.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, epriestley, mareksapota, davidreuss
Differential Revision: 1091
Summary: This should hopefully kill off the last of these :P
Test Plan: Should be self explanatory
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, epriestley
Differential Revision: 1043
Test Plan:
Used the scripts/mail_handler.php with and without patch and saw
the maniphest task being created with patch applied.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, epriestley, davidreuss
Differential Revision: 1041
transactional mail
Summary: See T571. SES refuses to deliver mail with this header and there are
various reports of other issues on the internet so I'm defaulting it to off.
Test Plan: Set config to true, tried to send mail, SES rejected it because of
"Precedence: bulk" header.
Reviewers: bmaurer, ola, jungejason, nh, aran
Reviewed By: aran
CC: aran, epriestley, bmaurer
Differential Revision: 1032
Summary: Allow configuration of a default author for bugs@ emails which don't
correspond to a known system user.
Test Plan: Configured a default author, sent some mails from nonsense addresses,
tasks were created.
Reviewers: davidreuss, jungejason, nh, tuomaspelkonen, aran
Reviewed By: aran
CC: aran, epriestley, ide
Differential Revision: 1013
Test Plan:
Set 'metamta.send-immediately' to true. Start up several MTA daemons, without
the patch you'll probably get multiple emails, with the patch you should get
only one.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, mareksapota, epriestley
Differential Revision: 1021
Summary: Quora wants to handle some moderation tasks with Phabricator, but want
to lower the barrier to entry for the install and let moderators adopt it
gradually. One request is to allow auth rules to be relaxed so we can auth based
on Reply-To to make things easier. This is insecure if configured but not really
a big deal and the patch isn't big or complicated.
Test Plan: Sent a test email with bogus "From" but valid "Reply-To". It was
rejected with this setting off, and allowed with this setting on.
Reviewers: jungejason, tuomaspelkonen, aran
Reviewed By: jungejason
CC: aran, jungejason
Differential Revision: 842
Summary:
This allows you to configure a single mailbox for all mail sent by phabricator,
so you
can keep a mailaddress like bugs@example.com and don't need a catchall on your
domain/subdomain.
Test Plan:
Enabled and disabled suffix. Saw mails generated have to correct prefix. Also
piped raw mails
into the scripts/mail/mail_handler.php and ensured comments went into
phabricator for both maniphest
and differential.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, epriestley
Differential Revision: 815
Summary:
@skrul reports receiving multiple copies of notification emails since
@hunterbridges configured some bizarre dystopian email replication factory on
their outbound route. Two fixes:
- Ensure "To" and "Cc" are unique. Email shouldn't be replicated for "To:
x@y.com, x@y.com" but it's silly that we do this.
- Remove "To" addresses from "Cc". Email shouldn't be replicated here either,
but we don't really lose anything by accommodating this.
Test Plan:
Sent a mail to the same to/cc, verified I was to'd only and not cc'd when the
mail was delivered.
@hunterbridges, can you apply this patch locally and verify it fixes the issue?
You can test by going to MetaMTA -> Send New Message and sending a message to
yourself as both To and CC.
Reviewed By: skrul
Reviewers: skrul, hunterbridges, jungejason, tuomaspelkonen, aran
CC: aran, skrul, epriestley
Differential Revision: 751
Summary: This lets you configure an email address which will create tasks when
emails are sent to it. It's pretty basic but should get us most of the way
there.
Test Plan: Configured an address and created a task via email. Replied to a task
via email to check that I didn't break that.
Reviewed By: tuomaspelkonen
Reviewers: davidreuss, jungejason, tuomaspelkonen, aran
CC: aran, epriestley, tuomaspelkonen
Differential Revision: 590
Summary:
We already support this (and Facebook uses it) but it is difficult to configure
and you have to write a bunch of code. Instead, provide a simple flag.
See the documentation changes for details, but when this flag is enabled we send
one email with a reply-to like "D2+public+23hf91fh19fh@phabricator.example.com".
Anyone can reply to this, and we figure out who they are based on their "From"
address instead of a unique hash. This is less secure, but a reasonable tradeoff
in many cases.
This also has the advantage over a naive implementation of at least doing object
hash validation.
@jungejason: I don't think this affects Facebook's implementation but this is an
area where we've had problems in the past, so watch out for it when you deploy.
Also note that you must set "metamta.public-replies" to true since Maniphest now
looks for that key specifically before going into public reply mode; it no
longer just tests for a public reply address being generateable (since it can
always generate one now).
Test Plan:
Swapped my local install in and out of public reply mode and commented on
objects. Got expected email behavior. Replied to public and private email
addresses.
Attacked public addresses by using them when the install was configured to
disallow them and by altering the hash and the from address. All this stuff was
rejected.
Reviewed By: jungejason
Reviewers: moskov, jungejason, tuomaspelkonen, aran
CC: aran, epriestley, moskov, jungejason
Differential Revision: 563
Summary:
See T251, where gregprice correctly argues that we need both:
None of the other people on the thread will have seen that message, so it
seems
like a lot of clients would put the server's message in a new thread. In
general, I think you want the References: header to mention every ancestor
message in the thread that you know about, because that's how MUAs keep a
thread
together in the face of missing some of its messages.
Test Plan:
Sent a reply email locally, got a response with both Message-IDs in
"references".
Reviewed By: rm
Reviewers: gregprice, rm
Commenters: gregprice
CC: aran, gregprice, epriestley, rm
Differential Revision: 499
Summary:
See T251. In Gmail, conversations split if you reply to them and the next email
does not "In-Reply-To" your message ID. When an action is triggered by an email,
carry its Message-ID through the stack and use it for "In-Reply-To" and
"References" on the subsequent message.
Test Plan:
Live-patched phabricator.com and replied to a Maniphest thread in Gmail without
disrupting the thread. Locally replied to Maniphest and Differential threads and
verified Message-ID was carried across the reply boundary.
Reviewed By: rm
Reviewers: tcook, jungejason, aran, tuomaspelkonen, rm
CC: aran, epriestley, rm
Differential Revision: 498
identifier
Summary:
Move the parser to a separate class so it can be easily unit tested, add some
tests. Properly parse emails with linebreaks in the quote line.
Test Plan:
Ran unit tests, used mail receiver to reply to an object.
Reviewed By: cadamo
Reviewers: aran, jungejason, tuomaspelkonen, cadamo
CC: aran, cadamo, epriestley
Differential Revision: 392
Summary:
Sendmail is seriously difficult to configure; SendGrid is extremely easy. It's
also pretty expensive ($80/mo) but there are a bunch of startups that already
have plans so it's effectively free for them.
Test Plan:
Configured SendGrid and sent reply email through it.
Reviewed By: aran
Reviewers: tuomaspelkonen, jungejason, aran
CC: aran, epriestley
Differential Revision: 376
Summary:
While my client and some others send email replies with an address like
##T1+x+y@example.com##, some other clients have sent either
##<T1+x+y@example.com>## or ##"T1+x+y@example.com" <T1+x+y@example.com>##.
Properly parse all the formats we've seen in the wild.
Test Plan:
Ran the regexp against all the formats observed in the wild (see
https://secure.phabricator.com/mail/received/) and verified it parses them
correctly.
Reviewed By: jungejason
Reviewers: tuomaspelkonen, jungejason, aran
CC: anjali, aran, jungejason
Differential Revision: 370
Summary:
There's an undoubtedly-far-more-refined version of this in xmail if someone
wants to crib it for me. Otherwise we can anneal this as counterexamples arise.
This seems to be what mail.app and gmail do.
Test Plan:
Used mail receiver console to "send" some mail and verified it was correctly
truncated.
Reviewed By: jungejason
Reviewers: aran, tuomaspelkonen, jungejason
CC: aran, jungejason
Differential Revision: 290
Phabricator
Summary:
Hook up the last pieces. This shouldn't impact the Facebook install, EXCEPT that
I removed "!accept" and added "!rethink" (plan changes). If you want to continue
supporting !accept, you should override the method in your subclass if you don't
already.
Test Plan:
Used the Mail Receiver test console to send mail to tasks and revisions.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran
Differential Revision: 289
Summary:
Provide an "isAdmin" flag for users, to designate administrative users.
Restore the account editing interface and allow it to set role flags and reset
passwords.
Provide an "isDisabled" flag for users and shut down all system access for them.
Test Plan:
Created "admin" and "disabled" users. Did administrative things with the admin
user. Tried to do stuff with the disabled user and was rebuffed. Tried to access
administrative interfaces with a normal non-admin user and was denied.
Reviewed By: aran
Reviewers: tuomaspelkonen, jungejason, aran
CC: ccheever, aran
Differential Revision: 278
Summary:
When we multiplex email, add information to the body with an explicit list of
recipients. Also add some headers if people want to write mail rules.
Test Plan:
Commented on a task and a revision, got reasonable looking emails about them.
Reviewed By: jungejason
Reviewers: aran, jungejason, tuomaspelkonen
CC: aran, jungejason, epriestley
Differential Revision: 272
Summary:
Provide a base PhabricatorMailReplyHandler class which handles the plumbing for
multiplexing email if necessary and supporting public and private reply handler
addressses. DifferentialReplyHandler now extends it, and a new
ManiphestReplyHandler also does.
The general approach here is that we have three supported cases:
- no reply handler, default config, same as what we're doing now
- public reply handler, requires overriding classes but just sets "reply-to"
to some address the install generates and still sends only one email
- private reply handler, provides a default generation mechanism or you can
override it and splits mail apart so we send one to each recipient
Test Plan:
Sent email from Maniphest and Differential with and without
reply-handler-domains set.
Reviewed By: aran
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran, epriestley
Differential Revision: 254
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
Summary:
Provides support for per-user x per-object unique reply-to email addresses, plus
SMTP integration.
This does not actually make Phabricator use these in outbound email.
Test Plan:
Used test console to validate in-Phabricator routing and handling.
Piped emails into the "mail_handler.php" script to validate mail parsing.
Configured sendmail and sent mail to Phabricator.
Technically I haven't conducted all parts of this test on the same machine since
I lost the will to configure more SMTP servers after configuring phabricator.com
Reviewed By: jungejason
Reviewers: jungejason, tuomaspelkonen, aran
CC: aran, epriestley, jungejason
Differential Revision: 226
Summary:
I pretty shortsightedly made sending a side effect of save() in the case that a
server is configured for immediate sending. Move this out, make it explicit, and
get rid of all the tangles surrounding it.
The web tool now ignores the server setting and only repsects the checkbox,
which makes far more sense.
Test Plan:
Sent mails from Maniphest, Differential, and the web console. Also ran all the
unit tests. Verified headers from Maniphest.
Reviewed By: rm
Reviewers: aran, rm
CC: tuomaspelkonen, rm, jungejason, aran
Differential Revision: 200
Summary:
Amazon SES does not allow us to set a Message-ID header, which means
that threads are incorrect in Mail.app (and presumably other applications
which respect In-Reply-To and References) because the initial email does not
have anything which attaches it to the rest of the thread. To fix this, never
rely on Message-ID if the mailer doesn't support Message-ID.
(In the Amazon SES case, Amazon generates its own Message-ID which we can't
know ahead of time).
I additionally used all the Lisk isolation from the other tests to make this
testable and wrote tests for it.
I also moved the idea of a thread ID lower in the stack and out of
DifferentialMail, which should not be responsible for implementation details.
NOTE: If you push this, it will cause a one-time break of threading for
everyone using Outlook since I've changed the seed for generating Thread-Index.
I feel like this is okay to avoid introducing more complexity here.
Test Plan:
Created and then updated a revision, messages delivered over Amazon
SES threaded correctly in Mail.app. Verified headers. Unit tests.
Reviewed By: rm
Reviewers: aran, tuomaspelkonen, jungejason, rm
Commenters: aran
CC: aran, rm, epriestley
Differential Revision: 195
Summary:
add a constants module
src/applications/phid/constants/PhabricatorPHIDConstants.
Test Plan:
Execute applications which were using the hard-coded string.
Differential Revision: 44
Reviewed By: epriestley
Reviewers: epriestley
CC: epriestley
Summary: Amazon SES seems to be working well, except that it takes more than a
second to send mail in-process. Kick it out of process. (Between this and the
ImplementationAdapter layer, MetaMTA almost makes sense. :/)
Test Plan: Ran the daemon and got a flood of unsent test email.
Reviewers:
CC:
