This commit adds code to pass configuration objects for group
permission settings in register response to clients such that
we do need to duplicate that data in clients and can avoid
future bugs due to inconsistency.
The "server_supported_permission_settings" field is included
in the response if "realm" is present in "fetch_event_types",
as this is what we do for other server-related fields.
This commit moves constants for system group names to a new
"SystemGroups" class so that we can use these group names
in multiple classes in models.py without worrying about the
order of defining them.
We now pass the complete configuration object for a setting to
access_user_group_for_setting instead of passing the configuration
object's fields as different variables.
This commit renames permissions_configuration variable to
permission_configuration since the object contains config for
a single permission setting and thus permission_configuration
seems like a better name.
See also 69e10b4dae.
Even though opts.msg_list can be undefined, we still want to retry
loading messages for it since our code needs that data regardless
of opts.msg_list being undefined.
If the user has no messages, `message_list_data` can be empty,
so, in that case, we just use the current oldest_message_timestamp.
The bug can also be reproduced if the last 400 messages were
in a muted topic for the user and we process recent_view_message_list_data
before all messages data.
We still process the remaining logic in `set_oldest_message_date` to
update has_found_newest/oldest.
This switches to our preferred dash-separated classnames
instead of underscore-separated, and also updates the
classname to be differentiable from users that aren't
part of the current narrow.
Until the next commit that splits the buddy list, the
name is a bit inaccurate since it still contains all
the users.
This was pvreviously covered by tests for the buddy list.
We shouldn't rely on other code to complete code coverage
for this module. In the future it would be good to add
more tests to the lazy set tests to directly cover lazy
set functionality.
This commit fixes the behavior of trying to remove the selected
text when selection is made from start in stream membership
input and compose box dm using backspace button.
Previously the selected text was not removed if no typeahead
was shown for the text and it instead removed the last pill.
Now we remove the selected text and not any input pills on
pressing the backspace button irrespective whether tyepeahead
is shown or not.
This commit updated the code to remove pill only when there is no
other text or when the pointer is at start of input but its selection
length is 0, in the input and let it follow the default behavior in
cases when there is any text (other than user pills) in the input
to fix the bug.
Updates the description of update_message_flags op: add event for
details about actions that send/trigger the event and other details
that are useful for client implementations.
Also, links to the above updates in the op: remove variant of the
update_message_flags event.
Previous behavior-
- Guest did not receive stream creation events for new
web-public streams.
- Guest did not receive peer_add and peer_remove events
for web-public and subscribed public streams.
This commit fixes the behavior to be -
- Guests now receive stream creation events for new
web-public streams.
- Guest now receive peer_add and peer_remove events for
web-public and subscribed public streams.
This commit updates code in bulk_remove_subscriptions and
bulk_add_subscriptions to return early if there are no
subscribers to remove or add to the streams.
This change helps us in avoiding unnecessary queries like the
one used to get subscribers list of streams, which is then used
to send events but we would not send any events if no subscribers
are added or removed and some more similar queries.
In ensure_customer_does_not_have_active_plan, we were already going
through the Customer table to get/check for an active CustomerPlan.
Now we directly get/check for an active CustomerPlan with via the
Customer, which allows for reusing this function for Customer
objects without a Realm set.
This is a case where the use of alpha channels in HSL color values
shows its limits.
The use of an alpha channel is necessary for the background
elements for the top navbar elements, because they sit over the
top of the navbar's bottom border (actually an inset shaddow).
However, it's impossible to use the alpha-channel based color on
elements like the unread dot, where the border actually sits on
top of the element itself--meaning that the effect would be a
larger dot with an imperceptibly darker ring around it.
What this commit does is use a technique suggested by Anders
Kaseorg for using CSS's `color-mix()` functional notation to
calculate an opaque version of the alpha color for use on
elements that do not or cannot directly take the color with
the alpha channel.
See CZO discussion:
https://chat.zulip.org/#narrow/stream/6-frontend/topic/alphas.20in.20color.20definitions/near/1670102
Moves two functions in corporate/lib/stripe.py that are used to
get data for the main installation activity analytics page to a
separate file: corporate/lib/analytics.py.
Also, updates these functions for the possibility of realm being
None for a Customer object.
`recent_view_focusable` class should be set on element whose
children can receive focus as per
`$topic_row.find(".recent_view_focusable").eq(col).children().trigger("focus")`
We should remove this overly defensive code and remove the `undefined`
type from the return type of this function to avoid handling unexpected
`undefined` values in the downstream code.
Now that the "new organization" form is quite tall, we were seeing
this autofocus logic causing the browser to automatically scroll to
the middle of the form (to reach the first editable input), which felt
like a pretty bad user experience.