Important changes in this commit:
* We only cache message list for "Combined feed" if it is the
default view.
* We modify existing handling of home message list code so that
it can be used to for any message list that we want to cache
using a new `preserve_rendered_state` variable.
* narrow_state.filter() returns the filter of combined feed view list
instead of `undefined`.
* We start fetching messages from the latest message on app load.
* Messages in all messages data and Recent view are always synced.
* If combined feed view list is not cached, we don't track it's
last pointer, effectively sending user to the latest unread
message always .
Previously, when a stream is deleted and the deleted stream is looked
up in the search bar, then an error is thrown.
This is because, when the StreamSubscription object of the deleted
stream is fetched we try to access its properties - invite_only and
is_web_public, despite it being undefined.
This is fixed by accessing the properties only when the object is not
undefined.
The update_selection function name was rather misleading, since that
function call is in fact what renders the message list object for the
view.
Also add comments about a few subtle/confusing details that I noticed
while debugging this code path today.
All messages list always has the "in-home" filter. We are in the
process of removing "zhome/zfilt" from message list, so this would
help in that process.
This was weird, and I think incorrect. Places that call add_message
seem to expect consistent data structures.
I tried looking a bit into it, and couldn't find anywhere where
returning true made more sense. I'm sort of confused this hasn't
caused issues though.
For more clarity, the 3 related functions `is_search` in `filter.js`,
`message_list.js` and `message_list_data.js` are all renamed to
`is_keyword_search`.
This effectively reverts a change introduced in #26260 that used
the original message to maintain the basline grid. With more
robust control over first-row line-heights, it is now safe to
fullyl hide the message again.
This also fixes a regression where messages with images would have
additional whitespace beneath the message edit box; see
https://chat.zulip.org/#narrow/stream/6-frontend/topic/message.20edit.20form.20height/near/1621758
Previously, the user setting "Mark messages as read on scroll" sounded
quite awkward and possibly confusing at times. We should update the text
to something more clear and concise. This change will change "Mark messages
as read on scroll" to "Automatically mark messages as read".
The help center article is also fixed to name which setting is being
adjusted.
This function will allow us to adjust the codebase to write what it
means semantically -- whether a check is for the message list being
visibly empty, or completely empty.
In this commit, we leave the .empty() method incorrect, because
several other adjustments need to be made atomically with fixing it.
There two situations where we might add messages to a previously empty
message list. The first is when a new message comes in; in that case,
the previous logic of just selecting the first message would work.
But in the case where we're instead adding a bunch of (previously
read) messages via a backfill API request, which can occur in a few
cases, including a stream narrow where all the topics with recent
messages are muted, we should use our standard rules to pick a message
to select, namely the first unread message (if any) or last message if
none are unread.
We've had a series of bugs where tooltips get leaked when a message list
is rerendered. For some tooltips, we used a 'mutation observer' to remove the tooltip
in this situation, but this was expensive and messy. We replace this with a Tippy
plugin to keep track of this class of tooltips, with a central hook to remove them
during rendering.
Message lists are rerendered in the background in a variety of situations;
a simple way to trigger it is clicking the mute/unmute topic/stream button in
the topic menu/stream menu and the clickable area overlaps with the
message list tooltips area. If a tooltip was visible at the time, the tooltip loses its
reference due to the re-rendering removing its DOM element, appearing at the top-left corner.
To prevent this behavior for all message list tooltips, we need to
store all instances of the message list tooltips and then destroy
them if the instances does refer to something else then document.body
using the 'destroy_all_message_list_instances' function just
before re-rendering.
Whenever the message list is rendered, all the message list tooltips
will be destroyed if they do not refer to document.body. This
prevents the double appearance of those tooltips if the reference
is removed from the DOM.
This plugin allows us to remove the mutation observers and net delete code
while hopefully fixing this bug for the whole app.
After merging #24309, we want to add an additional option to the "mark
messages as read on scroll" setting where we only mark messages as read
on scroll in conversation views.
We can dramtically simplify the stream muting live-update code path
for updating home_msg_list by observing the only thing it was doing
intentionally differently from update_muting_and_rerender is replacing
_all_items with a new version computed from from all_messages_data,
and the rest of update_muting_and_rerender can handle the live update
correctly unmodified.
This deduplication means live-update of "Unmute topic" just requires
updating the `in:home` filter logic appropriately, without any special
live update work.
These changes introduces an new user option where it allows
users read messages and scroll through their feed without
marking those messages as read.
Fixes: #22804.
Ever since we started bundling the app with webpack, there’s been less
and less overlap between our ‘static’ directory (files belonging to
the frontend app) and Django’s interpretation of the ‘static’
directory (files served directly to the web).
Split the app out to its own ‘web’ directory outside of ‘static’, and
remove all the custom collectstatic --ignore rules. This makes it
much clearer what’s actually being served to the web, and what’s being
bundled by webpack. It also shrinks the release tarball by 3%.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Aman Agrawal
roanster007
Tim Abbott
Karl Stolley
evykassirer
N-Shar-ma
evykassirer
lapaz
Joelute
PALASH BADERIA
Akarsh Jain
Anders Kaseorg