`postgresql-14.4` is a notable upgrade in the PostgreSQL series, as it
fixes potential database corruption from `CREATE INDEX CONCURRENTLY`
statements which are run while rows are modified[1]. However, it also
requires an upgrade from `libllvm9` to `libllvm10`, which means it is
not installed by a mere `apt-get upgrade`.
Add the `--with-new-pkgs` flag to all of the potentially relevant
`apt-get upgrade` calls, so that this (and similar) packages are
upgraded successfully.
[1]: https://www.postgresql.org/docs/release/14.4/
The images themselves had been deleted by commit
cc33b68d73, and were then zanitized out
of the commit history.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
30457ecd02 removed the `--mirror` from
initial clones, but did not add back `--bare`, which `--mirror`
implies. This leads to `/srv/zulip.git` having a working tree in it,
with a `/srv/zulip.git/.git` directory.
This is mostly harmless, and since the bug was recent, not worth
introducing additional complexity into the upgrade process to handle.
Calling `git clone --bare`, however, would clone the refs into
`refs/heads/`, not the `refs/remotes/origin/` we want. Instead, use
`git init --bare`, followed by `git remote add origin`. The remote
will be fetched by the usual `git fetch --all --prune` which is below.
While the `remote.origin.mirror` boolean being set is a very good
proxy for having been cloned with `--mirror`, is technically only used
when pushing into the remote[1]. What we care about is if fetches
from this remote will overwrite `refs/heads/`, or all of `refs/` --
the latter of which is most likely, from having run `git clone
--bare`.
Detect either of these fetch refspecs, and not the mirror flag. We
let the upgrade process error out if `remote.origin.fetch` is unset,
as that represents an unexpected state. We ignore failures to unset
the `remote.origin.mirror` flag, in case it is not set already.
[1]: https://git-scm.com/docs/git-config#Documentation/git-config.txt-remoteltnamegtmirror
The 0.1 second delay was sometimes not long enough to guarantee we hit
the async response path, resulting in a nondeterministic coverage
failure.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Commit 6fd1a558b7 (#21469) introduced an
await point where get_events_backend calls fetch_events in order to
switch threads. This opened the possibility that, in the window
between the connect_handler call in fetch_events and the old location
of this assignment in get_events_backend, an event could arrive,
causing ClientDescriptor.add_event to crash on missing
handler._request. Fix this by assigning handler._request earlier.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Adds a "Mobile" tab with straightforward instructions for accessing
the long-press menu using two new macros.
Knowing where to access the long-press menu is intuitive, except when
the user is in a topic narrow. It's not immediately obvious that the
top bar can be long-pressed so this adds a "Tip block" using a new macro
to clarify things in this scenario.
Also, this combines the instructions for Desktop and Web into a single
tab because the numbered steps work on both platforms. So this documents
the alternate method via the browser's address bar as a "Tip block" to
avoid stacking alternative numbered steps into a single tab.
This updates the stream and message link instructions too.
Fixes: #22147.
Removes the ":" which have accidentally ended up in the "Get a link
to a specific topic" and "Get a link to a specific stream" headings.
Renames the "Via browser's address bar" tab to "Web" so that it
stays consistent with other help center articles.
Fixes part of #22147.
Fixes#22182
This message often flashes on screen briefly, causing unnecessary
worry for the user (is the app likely to not load?).
To address this, we add a delay before the message is shown.
As a consequence, we change the notice to no longer suggest waiting a
few seconds, since we did that before showing it.
We increase the total number of messages, since increasing the number
of topics would otherwise have the side effect of making it hard to
find longer conversations.
The PR changes the following behaviors and UI:
1. Removes the checkmark button to mark the topic as read in
"Recent Topics".
2. Make the unread messages counter be the button for marking
all messages in the topic as read. The unread messages counter
is made clickable and tooltip is set to "Mark as read".
In "recent_topic_row.hbs", remove the checkmark button and add
classes and attributes to ".unread_counter" to give it desirable
behaviour on clicking.
In "zulip.css" set "opacity: 0.7" for ".on_hover_topic_read".
In "recent_topics.css" we set the background-color of unread counter to
hsl(105, 2%, 50%) to decrease fading of unread counter.
Fixes: #21654
Since `HttpResponse` is an inaccurate representation of the
monkey-patched response object returned by the Django test client, we
replace it with `_MonkeyPatchedWSGIResponse` as `TestHttpResponse`.
This replaces `HttpResponse` in zerver/tests, analytics/tests, coporate/tests,
zerver/lib/test_classes.py, and zerver/lib/test_helpers.py with
`TestHttpResponse`. Several files in zerver/tests are excluded
from this substitution.
This commit is auto-generated by a script, with manual adjustments on certain
files squashed into it.
This is a part of the django-stubs refactorings.
Signed-off-by: Zixuan James Li <p359101898@gmail.com>
We have now decided to not continue with the stream administrator
concept as we are changing the permissions model to be based on
user groups as per #19525. So, this commit updates the error message
to "Must be an organization administrator".
Adds minimal descriptions for email and general invitation links,
and documents that email invitations are translated into the
organization's notifications language.
94457732c1 changed this from:
```py
event_name = payload.get("event_name", payload.get("object_kind")).tame(check_string)
```
...to:
```py
event_name = payload.get("event_name", payload["object_kind"]).tame(check_string)
```
Which causes a failure when `event_name` exists but `object_kind` does
not, since the default is evaluated first.
Switch to an `if` statement to clarify the fallbacks better.
This function is oblivious to the existence of ArchivedAttachment, which
is incorrect. A file can be removed if and only if it is not referenced
by any Messages or ArchivedMessages.
Using http://localhost:9991 is incorrect - e.g. messages sent with file
urls constructed trigger do_claim_attachments to be called with empty
list in potential_path_ids.
realm.host should be used in all these places, like in the other tests
in the file.
We do not allow keeping vacant private streams as we deactivate them
when all users are unsubscribed, so it is better to add at least one
user in the initial private stream created while creating realm.
Without this, uwsgi does not release the GIL before going back into
`epoll_wait` to wait for the next request. This results in any
background threads languishing, unserviced.[1]
Practically, this results in Sentry background reporter threads timing
out when attempting to post results -- but only in situations with low
traffic, as in those significant time is spent in `epoll_wait`. This
is seen in logs as:
WARN [urllib3.connectionpool] Retrying (Retry(total=1, connect=None, read=None, redirect=None, status=None)) after connection broken by 'SSLError(SSLEOFError(8, 'EOF occurred in violation of protocol (_ssl.c:1131)'))': /api/123456789/envelope/
Or:
WARN [urllib3.connectionpool] Retrying (Retry(total=0, connect=None, read=None, redirect=None, status=None)) after connection broken by 'ProxyError('Cannot connect to proxy.', RemoteDisconnected('Remote end closed connection without response'))': /api/123456789/envelope/
Sentry attempts to detect this and warn, but due to startup ordering,
the warning is not printed without lazy-loading.
Enable threads, at a miniscule performance cost, in order to support
background workers like Sentry[2].
[1] https://github.com/unbit/uwsgi/issues/1141#issuecomment-169042767
[2] https://docs.sentry.io/clients/python/advanced/#a-note-on-uwsgi
Previously, when there were no unread messages in a topic, the topic
name was allowed to expand into the space allocated to the unread
count. This matched the behavior of the left sidebar. However, the
left sidebar has extremely limited horizontal space, and being able to
display a few extra characters is useful; recent topics does not have
this constraint. Further, recent topics wraps long topic names on
overflow, which looks ugly when using use the unread count's space.
So we switch to having the unread count element consume space even
when there is no count, using `visiblity:hidden`.
Fixes a part of #19449
Set the default_language as cookie and reload the page so that
the spectator can immediately see the language change in effect.
We can reload the page forcefully for spectators since there is
no chance of any work being lost. It is possible that the spectator
may lose the selected message on doing so.
This requires a new dependency, to be able to set cookies from
frontend JavaScript.
Fixes#21961
Alex Vandiver
Anders Kaseorg
Mateusz Mandera
Tim Abbott
David Rosa
Aman Agrawal
NerdyLucifer
Zixuan James Li
Alya Abbott
Sahil Batra
Lauryn Menard
evykassirer
madrix01