We had this API:
people.add_in_realm = full-fledged user
people.add = not necessarily in realm
Now the API is this:
people.add = full-fledged user
people._add_user = internal API for cross-realm bots
and deactivated users
I think in most of our tests the distinction between
people.add() and people.add_in_realm() was just an
accident of history and didn't reflect any real intention.
And if I had to guess the intention in 99% of the cases,
folks probably thought they were just creating ordinary,
active users in the current realm.
In places where the distinction was obviously important
(because a test failed), I deactivated the user via
`people.deactivate`.
For the 'basics' test in the people test suite, I clean
up the test setup for Isaac. Before this commit I was
adding him first as a non-realm user then as a full-fledged
user, but this was contrived and confusing, and we
didn't really need it for test coverage purposes.
Explicitly stubbing i18n in 48 different files
is mostly busy work at this point, and it doesn't
provide much signal, since often it's invoked
only to satisfy transitive dependencies.
We now treat util like a leaf module and
use "require" to import it everywhere it's used.
An earlier version of this commit moved
util into our "shared" library, but we
decided to wait on that. Once we're ready
to do that, we should only need to do a
simple search/replace on various
require/zrequire statements plus a small
tweak to one of the custom linter checks.
It turns out we don't really need util.js
for our most immediate code-sharing goal,
which is to reuse our markdown code on
mobile. There's a little bit of cleanup
still remaining to break the dependency,
but it's minor.
The util module still calls the global
blueslip module in one place, but that
code is about to be removed in the next
few commits.
I am pretty confident that once we start
sharing things like the typeahead code
more aggressively, we'll start having
dependencies on util. The module is barely
more than 300 lines long, so we'll probably
just move the whole thing into shared
rather than break it apart. Also, we
can continue to nibble away at the
cruftier parts of the module.
We just get the stream_name from the sub struct now.
This mostly affects node tests.
The only place in real code where we called add_sub()
was when we initialized data from the server.
When quoting a message with fenced code blocks without a language,
we used to have ambiguity in which '```' fence terminates the quote.
This commit adds explicitly non-interfering fences, which fixes the
above issue as well as makes the raw message easier to quickly read.
Fixes#12446.
This commit includes a new `stream_post_policy` setting,
by replacing the `is_announcement_only` field from the Stream model,
which is done by mirroring the structure of the existing
`create_stream_policy`.
It includes the necessary schema and database migrations to migrate
the is_announcement_only boolean field to stream_post_policy,
a smallPositiveInteger field similar to many other settings.
This change is done to allow organization administrators to restrict
new members from creating and posting to a stream. However, this does
not affect admins who are new members.
With many tweaks by tabbott to documentation under /help, etc.
Fixes#13616.
Zulip has had a small use of WebSockets (specifically, for the code
path of sending messages, via the webapp only) since ~2013. We
originally added this use of WebSockets in the hope that the latency
benefits of doing so would allow us to avoid implementing a markdown
local echo; they were not. Further, HTTP/2 may have eliminated the
latency difference we hoped to exploit by using WebSockets in any
case.
While we’d originally imagined using WebSockets for other endpoints,
there was never a good justification for moving more components to the
WebSockets system.
This WebSockets code path had a lot of downsides/complexity,
including:
* The messy hack involving constructing an emulated request object to
hook into doing Django requests.
* The `message_senders` queue processor system, which increases RAM
needs and must be provisioned independently from the rest of the
server).
* A duplicate check_send_receive_time Nagios test specific to
WebSockets.
* The requirement for users to have their firewalls/NATs allow
WebSocket connections, and a setting to disable them for networks
where WebSockets don’t work.
* Dependencies on the SockJS family of libraries, which has at times
been poorly maintained, and periodically throws random JavaScript
exceptions in our production environments without a deep enough
traceback to effectively investigate.
* A total of about 1600 lines of our code related to the feature.
* Increased load on the Tornado system, especially around a Zulip
server restart, and especially for large installations like
zulipchat.com, resulting in extra delay before messages can be sent
again.
As detailed in
https://github.com/zulip/zulip/pull/12862#issuecomment-536152397, it
appears that removing WebSockets moderately increases the time it
takes for the `send_message` API query to return from the server, but
does not significantly change the time between when a message is sent
and when it is received by clients. We don’t understand the reason
for that change (suggesting the possibility of a measurement error),
and even if it is a real change, we consider that potential small
latency regression to be acceptable.
If we later want WebSockets, we’ll likely want to just use Django
Channels.
Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
The compose_state.recipient field was only actually the recipient for
the message if it was a private_message_recipient (in the sense of
other code); we store the stream in compose_state.stream instead.
As a result, the name was quite confusing, resulting in the
possibility of problematic correctness bugs where code assumes this
field has a valid value for stream messages. Fix this by changing it
to compose_state.private_message_recipient for clarity.
This commit was originally automatically generated using `tools/lint
--only=eslint --fix`. It was then modified by tabbott to contain only
changes to a set of files that are unlikely to result in significant
merge conflicts with any open pull request, excluding about 20 files.
His plan is to merge the remaining changes with more precise care,
potentially involving merging parts of conflicting pull requests
before running the `eslint --fix` operation.
Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
We had a bug where if you started typing a message
and then used quote/reply (after the fact), we
would overwrite the user's original message.
The bug was kind of subtle--the internal call
to "respond" to the message would select the message
text, and then `smart_insert` would replace the
selection, unless it was Firefox.
Note that we now also allow you to cross-post
replies, which is a plausible scenario, although
possibly unintentional at times, too. I'm erring
on the side of giving the user control here, but
I'll add a warning in the next commit. Our compose
fade feature should also prevent unintentional
mixes here, too.
This seems like a small change (apart from all the
test changes), but it fundamentally changes how
the app finds "topic" on message objects. Now
all code that used to set "subject" now sets "topic"
on message-like objects. We convert incoming messages
to have topic, and we write to "topic" all the way up
to hitting the server (which now accepts "topic" on
incoming endpoints).
We fall back to subject as needed, but the code will
emit a warning that should be heeded--the "subject"
field is prone to becoming stale for things like
topic changes.
The stream/topic edit areas now have these ids:
#stream_message_recipient_stream
#stream_message_recipient_topic
They are pretty verbose, but being able to grep
for these without noise does have some value.
Use the placeholder `[Quoting…]` when quoting and replying before the
quote has been added to the message. Also, add tests to the
`compose_actions` Node tests for the new behavior.
Fix#10705.
This is part of work to break some of our
nastier circular dependencies in preparation
for our es6 migration.
This commit should facilitate loading leaf-like
modules such as people.js before all of the things
that reload.js depends on.
This run_test helper sets up a convention that allows
us to give really short tracebacks for errors, and
eventually we can have more control over running
individual tests. (The latter goal has some
complications, since we often intentionally leak
setup in tests.)
We flatten the code a bit by removing a check that type is object,
and we replace it later with a check that type is string.
We also no longer allow document-like objects to be wrapped based
on the location-attribute-is-present hack. Instead, we want the
tests to just set document to 'document-stub'.
Fixes#8965.
Mark_message(s)_as_read is used in marking a message as having been
read by the browser, rename it to notify_server_message(s)_read to
avoid any confusion.
@brockwhittaker wrote the original prototype for having
pills in the recipient box when users compose PMs (either
1:1 or huddle). The prototype was test deloyed on our
main realm for several weeks.
This commit includes all the original CSS and HTML from
the prototype.
After some things changed with the codebase after the initial
test deployment, I made the following changes:
* In prior commits I refactored out a module called
`user_pill.js` that implemented some common functions
against a more streamlined version of `input_pill.js`,
and this commit largely integrates with that.
* I made changes in a prior commit to handle Zephyr
semantics (emails don't get validated) and tested
this commit with zephyr.
* I fixed a reload bug by extracting code out to
`compose_pm_pill.js` and re-ordering some
calls to `initialize`.
There are still two flaws related to un-pill-ified text in the
input:
* We could be more aggressive about trying to pill-ify
emails when you blur or tab away.
* We only look at the pills when you send the message,
instead of complaining about the un-pill-ified text.
(Some folks may consider that a feature, but it's
probably surprising to others.)
Use compose_ui.insert_syntax_and_focus() when we need to insert text
inline-ly followed by the focus to compose textarea because it does
this job more smartly(it take cares of spaces).
The node tests have purged modules from cache that were
included via things like set_global(), but calling require
directly would leak modules into the next test, which made
a couple tests only work when you ran the whole suite. I
fixed those tests to work standalone. And then I now make
dependencies explicitly clear the require cache before we
require them in namespace.js.
This is mostly just moving methods out of compose.js.
The variable `is_composing_message`, which isn't a boolean, has
been renamed to `message_type`, and there are new functions
set_message_type() and get_message_type() that wrap it.
This commit removes some shims related to the global variable
`compose_state`; now, `compose_state` is a typical global
variable with a 1:1 relationship with the module by the same
name.
The new module has 100% line coverage, most of it coming
via the tests on compose_actions.js. (The methods here are
super simple, so it's a good thing that the tests are somewhat
integrated with a higher layer.)
This moves respond_to_mention() and reply_with_mention() to
compose_actions.js. These methods are basically thin layers
on top of compose_actions.start().
This module extracts these two functions that get called by
several other modules:
start()
cancel()
It is a little bit arbitrary which functions got pulled over
with them, but it's generally functions that would have only
been called via start/cancel.
There are two goals for splitting out this code. The first
goal is simply to make `compose.js` have fewer responsibilities.
The second goal is to help break up circular dependencies.
The extraction of this module does more to clarify
dependencies than actually break them. The methods start()
and cancel() had actually been shimmed in an earlier commit,
and now they no longer have a shim.
Besides start/cancel, most of the functions here are only
exported to facilitate test stubbing. An exception is
decorate_stream_bar(), which is currently called from
ui_init.js. We probably should move the "blur" handler out
of there, but cleaning up ui_init.js is a project for another
day.
It may seem slightly odd that this commit doesn't pull over
finish() into this module, but finish() would bring in the
whole send-message codepath. You can think of it like this:
* compose_actions basically just populates the compose box
* compose.finish() makes the compose box do its real job,
which is to send a message
Steve Howell
Vishnu KS
Rohitt Vashishtha
Ryan Rehman
Anders Kaseorg
Tim Abbott
YashRE42
Mohit Gupta
Marco Burstein
Shubham Padia
Shubham Dhama
cPhost
Tommy Ip
Joshua Pan
Harshit Bansal
Steve Howell
Mahim Goyal