Setting `credential_source` is used when assuming role credentials --
that is, when running as one role, use the AssumeRole right to become
someone else.
The AWS command-line tools only do this if `role_arn`, the role to
assume, is also set -- if it is not set, it transparently falls
through to IAM role attached to the EC2 instance profile. However,
with the `aws-sdk-go` package, used by Teleport, this configuration
produces an error.
Remove the `credential_source = Ec2InstanceMetadata` line, which isn't
necessary for the AWS CLI, and interferes with Teleport operation.
This commit updates the urls for personal narrow sent in email
notifications to be of form "{user_id}-{encoded_full_name}" to
make it consistent with the urls that we use for such narrows
in webapp which were recently updated in b4eddad for improving
performance. We encode the full name in the same way that we do in
webapp by replacing the url characters encoded by browser with "-".
It doesn't seem to make sense to append _{number} to the status code in
that arg, because the resulting string stops looking like a status code
and actually makes this test fail in the follow-up commit with the
confusing error message of
Unknown response http status: 2000
So this just seems like a bug.
Previously, when you clicked the confirmation button in a
dialog_widget, we'd display a loading spinner on top of an ugly grey
button background.
Remove the ugly grey background color, while making sure the loading
spinner inside the button looks good and doesn't resize the button.
Fixes: #22002
This change is made so that we are not accessing the global page_params
directly. Note that we also rename unread_msgs to params to match other
modules that follow this pattern.
Signed-off-by: Zixuan James Li <p359101898@gmail.com>
This adds a get_unread_message_count helper that returns the size of
unread_messages and replaces the usage of page_param.unread_msgs.count
elsewhere in the codebase.
Fixes#23334.
Signed-off-by: Zixuan James Li <p359101898@gmail.com>
This setting provides a more ergonomic experience when viewing a
message where only a small number of emoji reactions are present; you
no longer need to hover over the reaction to see who reacted, which is
often quite important to understanding what it means.
We added this setting (visible in the development environment only,
and implemented on mobile) a few months ago; so to complete this
feature, we just need to add the frontend rendering/live update logic
and publish it.
As documented in the code comments, the current rule is "messages with
3 or fewer reactions" for which messages will display the names of the
users reacting.
The previous set_reaction_count API made sense when the display of a
given reaction was a pure function of its data. However, we will soon
be making it a function of the total number of reactions on the
message.
Thus, view functions to update reaction display now need to call a
shared update_vote_text_on_message function in all code paths, even
code paths like "add the first vote for a reaction" and "remove the
sole vote of a reaction" which previously did not need to call
set_reaction_count.
Fixes#20890.
We may want to rename clean_reactions to something else, but the name
`r` is pretty confusing for an object that isn't a single reaction,
but instead an object detailing all the reactions on a message.
The blue font color on the vote_text beside the reactions in
light mode made it seem like a link which it's not, so it has
been changed to regular font color.
We preserve the blue font color on hover, matching the add-reaction
button.
This fixes a bug where in narrow windows, long vote_text could extend
beyond the body of the reaction's allocated area.
It also provides a matching right margin around the count/name inside
the reaction pill, which looks better when we have strings inside them.
Follow-up to commit 22a5d008c1.
The new display setting introduced in above commit was not registered
properly and enabling/disabling it from Organization settings > Default
user settings did not display a "Save/Discard" widget. This has been
fixed by modifying the `settings_org.get_subsection_property_elements`
function.
Since stream header is now sticky, we need to reduce it's height
when calculation position of the element we want to scroll into view
so that it doesn't hide under the sticky header.
This commit introduces the change of rendering private messages
section as collapsible, whose data-fetching logic came with zulip#21357.
We now have separated out `Private messages` from `top_left_corner`
section and shifted it below the `global_filters` in a different
separate section along with stream list with common scroll bar
in left-sidebar.
The new PM section will be opened by-default on loading the page
and will have a toggle-icon in its header, clicking on which makes the
section collapse/expand accordingly.
In default view, only recent 5 PM threads would be shown
and would append the active conversation as the 6th one at last
if not present in those 5, similar to how topics list work.
In PM section with unreads, a maximum of 8 conversations
would be shown and rest of them would be hidden behind
the 'more conversations' li-item, clicking on which takes
to the zoomedIn view of PM section where all the present
PM threads would be visible and rest of the sections of left-sidebar
will get collapsed.
Fixes#20870.
Co-authored-by: Aman Agrawal <amanagr@zulip.com>
Dropped text "(Message sent when you were not subscribed)" for
historical messages in the three-dot message menu.
This information was already available via the "You subscribed" type
divider lines, which were added about the same time back in 2013, and
this text feels somewhat ugly.
Also removed the historical field from the args object that is
passed to the template, as this is not used anymore.
Fixes#23198.
Some options in the profile menu are not documented in the help center.
Documents "View messages sent".
Documents "View messages with yourself".
Links `sender:me` search operator to the relevant section on this page.
Fixes: #23203.
Updates the `filter.get_title` logic to return the stream name for
narrows that include the stream, topic and near operators. That
way the browser/tab title remains the same for these views, which
have a particular scroll offset.
Change URLs on "Choosing between Zulip Cloud and self-hosting" page
to be non-relative when pointing to the Zulip landing page or `/new`
for creating a new Zulip Cloud organization.
Removes call to reset_compose_message_max_height when clicking on
the markdown preview button, which due to the `#compose` div element
momentarily shrinking to be empty, caused the calculation of the
max-height to grow larger on each click.
Also refactors reset_compose_message_max_height to use the height
from `getBoundingClientRect`, which defaults to zero when empty.
And fixes a small discrepancy in how max-height is applied to
a div element vs a textarea element, so that the visible height
doesn't change between the preview and write modes in the compose
box.
Fixes#23277.
Before Zulip 4.9, the Zulip install process left any already-installed
rabbitmq with whatever nodename it had previously configured. Wince
this encodes the name of the host when it was installed, this does not
function well with containers.
Leave rabbitmq-server uninstalled, which lets the Zulip installation
process set the nodename to `localhost`, which ensures that it is
usable across container restarts.
This gives us some time to render and fetch data before allowing
user to start using hotkeys. It avoids error being thrown when
hotkeys are used when the app is still loading.
Updates the hash used for the recent conversations view to be
"#recent" instead of "#recent_topics".
We will need to keep the logic for handling "#recent_topics"
permanently because users potentially have messages from
Welcome Bot with links to that hash.
Including "recent_topics" as a web_public_allowed_hash in
hash_util.js can be changed once self-hosted servers cannot
upgrade directly to Zulip 5.x from the current version.
Fixes#23132.
The resize handler is a somewhat expensive operation, which we'd like
to avoid when processing the receipt of new messages. It's also
unclear why it would be necesssary; bottom_whitespace is fixed in
size, and the sidebars.
It should be very rare to discover new unread messages during a
message_fetch call. This can potentially happen due to races (fetching
just as a new message arrives), but it shouldn't be the common case.
Previously, we would trigger a full rerender of all UI displaying
unread messages every time a bulk message fetch operation returned
(including every time one narrowed), regardless of whether any actual
state had changed.
Fix this by actually checking if we discovered any new unread messages.
There's no reason that opening left sidebar popovers should require
the app to recalculate the size of any of its components; popovers are
drawn on top of the rest of the UI in any case.
This appears to have been added in
9ed05a1f85 as part of prototyping the
left side user list feature.
This was likely introduced in error in
fd66d9f70336b2bdbbffb1bf13402516b9920682; it is a duplicate of
show_userlist_sidebar, which, in contrast is actually called.
Because we can count on the unread.js data structures to be correct,
we no longer need this legacy code that tries to re-process all
messages the client has cached for whether any are present in the
target stream and marked as unread.
It's not clear we need this at all, since the current Zulip server
will mark all messages as read when unsubscribing from a stream, but
having this call protects us from future bugs should that behavior
change.
This cleanup also revealed a possible bug, where we would fail to
update global unread counts properly when unsubscribing from a stream
until the deferred_work queue processor sends us the
update_message_flags event associated with that change.
The previous invocation message_util.do_unread_count_updates dates
from a time when `unread.js` did have any data beyond what full
message objects were present in the client, and thus we needed to
recalculate everything from those message objects when muting state
was changed.
With the modern unread.js data structures, all that's required is a
full recalculation/rerender via this function.
We change the "pm-with" and "sender" narrow urls
to be of "{user-id}-{encoded_name}" form instead
of using email. This change improves performance
of changing between PM views since parseOneAddress
function was slow and we remove its usage now by
using name instead of email.
The name is encoded such that the characters that
would be encoded by the browsers are replaced by
"-".
Alex Vandiver
Alya Abbott
Sahil Batra
Mateusz Mandera
aparna-bhatt
Zixuan James Li
Tim Abbott
Raghav Luthra
Aman Agrawal
jai2201
Barissa-Imran
Rebecca Södereng
David Rosa
Lauryn Menard
Matt Keller
Rishant Rokaha