`setup_path()` previously only checked that some `zulip-py3-venv` was
the `sys.prefix`, not that it was the one associated with this
deployment. When `uwsgi` is started, it is started from `bin/uwsgi`
within a `zulip-py3-venv` virtualenv, and as such sets
`sys.executable` to that, resulting in uwsgi workers picking up the
library path of that virtualenv. On first start, `sys.path` thus
already matches the expected virtualenv, and the `setup_path` in
`zproject.wsgi` does nothing.
If a rolling restart was later done into a deployment with a different
virtualenv, the `zproject.wsgi` call to `setup_path()` did not change
`sys.path` to the new virtualenv, since it was already running within
_a_ virtualenv. This led to dependency version mismatches, and
potentially even more disastrous consequences if the old (but still
erroneously in use) virtualenv was later garbage-collected.
PR #26771 was a previous attempt to resolve this, but failed due to
not thinking of the uwsgi binary itself as possibly providing a
virtualenv path. We leave the `chdir` hooks from that in-place, since
it cannot hurt for the "master" uwsgi process to be chdir'd to `/`,
and the `hook-post-fork` `chdir` is reasonable as well.
Resolve the virtualenv in `setup_path()`, and activate it if it
differs from the one that is currently active. To be sure that no
other old virtualenvs are used, we also filter out any paths which
appear to be from other Zulip virtualenvs.
We should not proceed and send client reload events until we know that
all of the server processes have updated to the latest version, or
they may reload into the old server version if they hit a Django
worker which has not yet restarted.
Because the logic controlling the number of workers is mildly complex,
and lives in Puppet, use the `uwsgi` Python bindings to know when the
process being reloaded is the last one, and use that to write out a
file signifying the success of the chain reload. `restart-server`
awaits the creation of this file before proceeding.
Thumbnails are usually enqueued in the worker when the image is
uploaded. However, for images which were uploaded before the
existence of the thumbnailing worker, and whose metadata was
backfilled (see previous commit) this leaves a permanent spinner,
since nothing triggers the thumbnail worker for them.
Enqueue a thumbnail worker for every spinner which we render into
Markdown. This ensures that _something_ is attempting to resolve the
spinner which the user sees. In the case of freshly-uploaded images
which are still in the queue, this results in a duplicate entry in the
thumbnailing queue -- this is harmless, since the worker determines
that all of the thumbnails we need have already been generated, and it
does no further work. However, in the case of historical uploads, it
properly kicks off the thumbnailing process and results in a
subsequent message update to include the freshly-generated thumbnail.
While specifically useful for backfilled uploads, this is also
generally a good safety step for a good user experience, as it also
prevents dropped events in the queue from unknown causes from leaving
perpetual spinners in the message feed.
Because `get_user_upload_previews` is potentially called twice for
every message with spinners (see 6f20c15ae9), we add an additional
flag to `get_user_upload_previews` to suppress a _second_ event from
being enqueued for every spinner generated.
We previously used the file extension to determine if we should
attempt to inline an image. After b42863be4b, we rely on the
existence of ImageAttachment rows to determine if something is an
image which can be viewed inline. This means that messages
containing files uploaded before that commit, when (re-)rendered, will
be judged as not having inline'able images.
Backfill all of the ImageAttachment rows for image-like file
extensions. We are careful to only download the bytes that we need in
the image headers, to minimize bandwidth from S3 in the event that the
S3 backend is in use. We do _not_ produce thumbnails for the images
during this migration; see the subsequent commit.
Because this migration will be backported to 9.x, it is marked as only
depending on the last migration in `9.x`, with a subsequent merge
migration into the tip of `main`.
Expands the header information on the plan activity view to include
the license ledger that was set as invoiced_through when the last
billing cycle generated an invoice.
Also, updates the name of the header information for the next
invoice date to match what's show in the support view for the plan.
Includes the license counts, date and renewal information as well
as the id field, so that print debugging and displaying these
objects is easier and more useful.
Display "No channels match your filter" in the left panel
when no matches are found for filter queries in
"Subscribed", "All channels" and "Not subscribed" tabs.
Fixes#18332.
Adds a link to the plan ledger view in the current plan information
shown in the support views. Link is not shown if the plan is 100%
sponsored, e.g., the Community plan.
Adds a formatted header area to the activity table template so
that it's easy to add useful information to these activity views.
Since each loop may add more than one file to the `storage_paths`
list, this may result in more than 1000 files being sent to
delete_message_attachments. Since the S3 backend only supports 1000
elements being deleted at once, we must partition the list into chunks
which are no more than 1000 elements long.
Previously the empty view banner for DM with any recipients
contained "Why not start the conversation".
For deactivated users/bots, we are now not displaying it.
The "Role" field previously only displayed "Bot" for the non-system
bot users, this commit adds the detailed bot role to the "Role" field
in the format "Bot ({role})".
This commit displays the bot owner card as an overlay in the middle of
the screen, like we do for mobile screen sizes where the reference
elements are hidden.
This also hides the bot user card when it's bot owner card is opened.
As a follow-up to the previous commit which repositions the bot owner
field, this commit adds the bot's role in the format "Bot ({role})" to
the previous position of the bot owner field in the bot user card.
This commit repositions the bot owner field from below the bot’s name to
align with the user fields, providing more space for the newly added
bot owner user pill.
Fixes the URLRedirects for "help/about-streams-and-topics" and
"help/streams-and-topics" to go to "/help/introduction-to-topics"
since "help/channels-and-topics" has been redirected to that URL.
Earlier when compose box was opened or narrowed to pm, the recipient
pills were sorted according to user email strings instead of their
full names which was inconsistent with compose box placeholder text,
recipient row and message header.
This commit fixes the behaviour by introducing a `sort_email`
function to sort emails according to their full names and display
sorted pills.
Fixes: zulip#27375.
Co-authored-by: richardshaju <richardshaju66@gmail.com>
Earlier in drafts overlay, the order of pm recipients for the draft
was sorted differently to that of the pm group title.
This commit fixes the behaviour by sorting the drafts pm recipients
using `make_strcmp`.
Fixes part of zulip#27375.
Co-authored-by: richardshaju <richardshaju66@gmail.com>
Earlier in left sidebar, the recipient names were not sorted
consistently with the message header of the pm.
This commit fixes the behaviour by sorting the recipient names in
left sidebar with `make_strcmp`.
Fixes part of zulip#27375.
Co-authored-by: richardshaju <richardshaju66@gmail.com>
Earlier, in group direct message, the full names of group pm recipients
were sorted incorrectly which is used for tooltip on hovering over
recipient row and text placeholder of collapsed compose box. That
resulted in different order of recipient names.
This commit fixes the behaviour by using the `make_strcmp` to sort
the names to show in the tooltip and collapsed compose box.
Fixes part of zulip#27375.
Co-authored-by: richardshaju <richardshaju66@gmail.com>
Earlier, when narrowed to a group direct message, the header title
was sorted with respect to its email instead of their names. This
would result in message header and recipient row having different
order of recipients.
This commit fixes the behaviour by changing the order of the names
passed for the group message header using `make_strcmp`.
Fixes part of zulip#27375.
Earlier, the order of recipients in recipient row of group direct
message weren't in the same sorting pattern as that of message
title which would result in inconsistent order of recipient names.
This commit fixes the behaviour in `private_message_header` in
recipient row to sort the names using `strcmp`.
Fixes part of zulip#27375.
Co-authored-by: richardshaju <richardshaju66@gmail.com>
This also moves the 2px of reduced lefthand padding to the overall
width of the left sidebar, which will keep the left sidebar at its
same width (in other words, the message area will not shift any
further to the left, either).
Fixes#27565
Adds non-form section to Zulip Cloud support view with some basic
realm information: organization type, plan type, non-guest user
count and guest user count.
Uses a shared template for the basic realm data and adds a shared
support context dict for variables that are used in both remote
and Zulip Cloud support views.
Gives a baseline of current database queries for these single realm
or remote realm support views, so that as we add features to these
views, we can better manage how those changes impact the performance
of our support views in general.
There are a few places where we want to set the max invites for a
realm to the default for a realm's plan type, so this creates a
helper function that can be used consistently to get that default
value.
This is part of a bigger refactor to calculate message container
attributes just from the message, so that we can create the message
container all at once instead of piecewise, to be able to convert
to typescript.
