The code removed in this commit added pinned streams to a list
of elems, only to have them added again by the next block of code
(but more concisely). Through some strange quirk of appendTo() this
never created user-facing bugs, but you could clearly see in the console
that it was doing double work.
It used to be the case that you would get new messages for a
huddle, but the huddle wouldn't show up on your buddy list until
the every-50-seconds mass update of the buddy list.
Now we make sure to work with non-stale jQuery objects, and,
more importantly, we resize ourselves if we add new huddles.
(The resize issue arises due to some complicated heuristics
where we don't want group PMs to take up too much of the buddy
list for users who don't have many in their history.)
This fixes a bug with the group pm section of our
buddy list. It wasn't updating when you fetched
old private messages.
We had been calling activity.process_loaded_messages() as
part of message_store.do_unread_count_updates(), which was
called sometimes unnecessarily and sometime not called when
we needed to get huddle info.
Now we call it when we need it most, which is when you
click on "Private Messages".
Previously, if a new message arrived between when a user is subscribed
to the default streams and when the user's initial messages are
queried, we would try to create two UserMessage rows for the same
Message, resulting in an IntegrityError crash. We fix this and add a
test for that race condition.
This restyles the message_controls options to center them horizontally
while fixing them closer to the right side of the edge, along with just
replacing the edit button with a preview source button once editing is
disabled.
This moves the edit button to underneath the timestamp such that when
you hover over a message now the timestamp hides itself and the edit
button appears (if editing is allowed).
Fixes#1733 and other annoying issues with this field.
* Fixes handling of multiple stream links and invalid stream names.
* Fixes text regex so it handle hash sign the right way.
* Adds tests for these stream link cases.
Changes editibility type from NONE to FULL, and makes Save and Cancel
buttons work. Exposes a new bug, which is that changing the topic changes
the topic for the recipient row, rather than the message. See #2278.
The widget that gets built in topic_list.build_widget()
now knows how to add itself to its parent element and expose
an interface to retrieve the parent.
For cases where we are zoomed in to a stream and then go
to a different narrow (Home, PMs, etc.), we now let
topic_list.zoom_out orchestrate the removal of the topic list
instead of stream_list.zoom_out.
This will help us when we move to a world where topic_list
redraws topic lists on zoom-in/zoom-out, because we won't
waste effort rebuilding lists that are about to be removed.
The flow for topic list zooming is kind of complicated now, but
it's mostly a consequence of the way the UI works.
* stream_list tells topic_list to set up the topic list
click handlers to have callbacks to stream_list
* topic_list click handlers call to stream_list zoom methods
to hide/show all the other streams
* stream_list zoom methods call back to topic_list methods to
redraw topics as needed (this isn't happening yet, but allowing
topic_list.js to know that it's zoomed will set the stage for
this to happen in a more controlled manner)
We no longer use active_stream_name(), which was mostly a
duplicate of narrow.stream(). For nonsensical queries like
"stream:foo stream:bar" the behavior may change slightly here.
We know that we don't handle non-sensical queries particularly
well, but at least if we always go through narrow.stream(),
the behavior will be consistent.
I did test this with some sensible compound narrows, like searching
for a keyword within a stream.
The filter args dictionary applies to the X table in a count X by Y query,
which in this case is the zerver_message table. This stat had an incorrect set
of arguments meant for the zerver_userprofile table.
This migration ensures that all historically uploaded files from
before we started tracking files in the Zulip database via the
Attachment model have Attachment objects.
This has been tested by tabbott against a production server to ensure
that it results in all old uploaded files having corresponding
attachment objects.
Merging this change is a key prerequisite for making our adding
attachment access controls in an enforcing fashion.
In Django 1.10, the get_token function returns a salted version of
csrf token which changes whenever get_token is called. This gives
us wrong result when we compare the state after returning from
Google authentication servers. The solution is to unsalt the token
and use that token to find the HMAC so that we get the same value
as long as t he token is same.
In 25b28bf82c and then
cb1bc70ab0, we attempted to make long
code blocks scroll in a reasonable fashion, without much success.
This change causes code blocks to be line-wrapped, without needing to
set `overflow-y: hidden` for paragraphs (which cause problems with
taller elements like emoji that overlfowed wrong). Our octopi finally
have legs again.
It's not clear that this is the final answer, but it's the best
version we've found so far.
Needed in case the user was allowed to join the realm when they got the
confirmation email, but is no longer allowed to do so. Check was previously
applied to invited users (those with a prereg_user.referred_by), and is now
applied regardless of how they get to accounts_register.
Steve Howell
Tim Abbott
Brock Whittaker
sinwar
paxapy
Igor Tokarev
Rishi Gupta
umkay
Umair Khan