This commit removes bot_error function in settings_bots.js
since it is not used now. It was added in d90d1a04 and was
used in error handling, but since we changed the design to
show dialog for reactivation and deactivation of bots and
errors are shown inside the dialog and errors are handled
differently and thus the function's usage was removed in
a9893fb654.
This commit also removes get_bot_info_div which was only
used by bot_error function.
We already have "inline-block" class for the h3 element
in profile_settings.hbs, so the display property is set
as "inline-block" and the CSS set by "hide" class is
overridden. We should not have "hide" class for this
element, since we want to show that element and not
hide it.
This was probably added while picking code of some
other element in bb816e199.
Removes the `btn-direct` class in `portico.css` that was only
being used for dev login buttons.
Adds `dev-button` class for general CSS rules for buttons on the
dev login page. Adds `dev-login-button` and `dev-create-button`
classes for CSS rules specific to the two types of buttons on the
page.
Removes `#find_account .btn` and `#find_account .form-control`
rules in `static/styles/portico/portico.css`.
The last use of these rules was removed in commit 7afbc9ddd6
when the login and registration pages were redesigned.
This commit moves the upgrade text for message retention
setting below the label so that it is clear that the text
is present for message retention setting. This change is
done for both stream creation form and stream edit panel.
The upgrade text shown for message retention setting for realms
on limited plans should mention about sponsorship only for
non-business orgs. In the stream edit panel, the sponsorship
text was present even for non-business orgs because the
is_business_type_org parameter was not being passed to the
template, so this commit fixes the code to pass the
is_busines_type_org parameter correctly to the template.
The message retention setting is only visible to owners in the
stream creation form, so the jquery code to hide the custom
input, set the default value for dropdown and listener to
show and hide the custom input should be called only for owners.
The `next_is_same_sender` has no effect on the CSS of the message
displayed and the JS changes seem to have no effect too.
See cc8021a742 for more details.
The English word "at" was manually appended to the string output of
datetime-related functions to generate the string shown in the tooltip
when hovering over the timestamp of a message. Use the translated form
"{date} at {time}" instead, as found elsewhere in the codebase.
Disables the deactivate account button in the user's account and
privacy settings tab if they are the only active organization owner.
Adds a tooltip when hovering on the deactivated button to let the
user know why the button is disabled.
The backend already returns an error for self account deactivation
requests if the user is the only organization owner.
Adds `avatar.build_user_avatar_widget` to the updates done in
`settings_account.update_avatar_change_display` when the user
has permission to change/delete their avatar.
For the case where a user has their personal account profile
settings open while the organization or user's permissions are
changed, `user_avatar_widget_created` boolean is added to
`settings_account.js` to track whether the widget and handlers
have been created so that live updates don't rebuild the
existing widget and handlers.
Also, updates `avatar.build_user_avatar_widget` to return early if
the user cannot change their avatar. This pattern better matches
the other instances where we use the direct upload widget for the
organization icon and logos.
The "day" and "night" color scheme names were updated in #20371,
but the organization profile settings tab still has had the old
names for the two types of wide logos.
Updates "Day logo" to be "Light theme logo" and "Night logo" to
be "Dark theme logo".
Adds hiding any elements with the `image_upload_button` class in
the organization profile (icon and logos) to the live updates in
`settings_org.maybe_disable_widgets`, so that any role changes
that remove admin permissions will update these images so that
they no longer appear able to be updated / deleted.
The `settings-info-icon` with the tooltip for the user avatar
section was removed in commit 103db2afaf. This removes the changes
to that tooltip in `settings_account.update_avatar_change_display`.
Instead of hiding or showing the `image_upload_button` in that same
live update function, we now add or remove the `hide` class, which
in the previous commit has a specific CSS hover rule for the image
upload widget.
Also adds test coverage for the `is_admin` case for this setting
because administrators can always change their own avatar even if
the organization has disabled avatar changes.
If there is no `disabled_text` provided for the image upload widget,
then there is no need to have the `image_disabled` div rendered in
the template. This allows the hover CSS rules to be more general for
the image upload widget in general.
Adds a check for `disabled_text` around the `image_disabled` div
element in `image_upload_widget.hbs`.
Also, changes `image_upload_background` class to
`image_hover_background` so that it more accurately describes what
the CSS rule is for and why it's used in both the `image_disabled`
div and the `image_upload_button` div.
The `hide` CSS rule in `app_components.css` was being overwritten
by the more specific rules in `image_upload_widget.css`, which
meant that when changing the inage was disabled the hover text for
updating or deleting the image was still visible.
Adds `hide` class (and therefore more specific rule) to
`image_upload_widget.css` for when this text should be display
none.
Fixes#23844.
This only set the last_message CSS class, which is no longer used for
styling or JavaScript code.
(The calculation was also wrong, in that new messages arriving would
not cause it to be removed from the previous message with the
last_message class).
The last_message CSS class didn't mean what it said it did, due to
issues with live update.
Further, this logic was poorly written, with `$message` changing types
from a .message_row to a .recipient_row for now apparent reason.
I was able to reproduce at least one bug where the `v` shortcut would
not correctly open the lightbox that is fixed by this rewrite.
Fixes zulip#19759
Before grid, there was overlap of action buttons and text (photo 1).
With grid, there was just very little space for messages (photo 2).
With this commit, the message gets a bit more space (photo 3).
This helps us add more width to the content column of message
on narrow widths. Note that it will still always be less than
before grid, since we are explicitly preventing overlap.
Using grid layout moves us away from fixed widths
and position for items where possible, which will
make it easier to make messagebox changes without
breaking formatting.
Visual changes expected in this commit:
- When the action buttons overlap the message content,
it will no longer visually overlap but completely cut off
the content — which is good, but means the messages can’t
be as wide on narrow width views. There’s a slight improvement
to this in an upcoming commit.
Also removes the `last_message` CSS, which was busted due to the
`last_message` calculation not correctly being updated for new
messages arriving, and didn't improve styling.
Having the class at the top of message DOM structure when the
user is editing a message, helps apply css when user is editing
a message.
Short prep commit for the next commit; no visible changes
Previously, we got the directory path for all documentation pages
before checking for API method and path information in the OpenAPI
documentation. Instead, we now check the `path_template` is the
API documentation view template before getting the directory path.
Also, changes the confusingly named `article_path` variable, which
overlapped with the DocumentationArticle dataclass `article_path`
field, to now be `api_documentation_path`.
Prep commit for moving the help center documentation to a top level
directory.
Accessing .realm will cause a fetch query from the database if the
attribute hasn't been fetched already earlier in the codepath. That's
completely redundant if we're just comparing realms, and we should only
access .realm_id attribute. This seems to eliminate a query in some
codepaths, which is nice in this performance-sensitive function.
Adds links to the documentation about management commands in the
API documentation for creating users, as well as the `/devtools`
documentation, the GDPR compliance article and the incoming
webhooks tutorial.
When file uploads are stored in S3, this means that Zulip serves as a
302 to S3. Because browsers do not cache redirects, this means that
no image contents can be cached -- and upon every page load or reload,
every recently-posted image must be re-fetched. This incurs extra
load on the Zulip server, as well as potentially excessive bandwidth
usage from S3, and on the client's connection.
Switch to fetching the content from S3 in nginx, and serving the
content from nginx. These have `Cache-control: private, immutable`
headers set on the response, allowing browsers to cache them locally.
Because nginx fetching from S3 can be slow, and requests for uploads
will generally be bunched around when a message containing them are
first posted, we instruct nginx to cache the contents locally. This
is safe because uploaded file contents are immutable; access control
is still mediated by Django. The nginx cache key is the URL without
query parameters, as those parameters include a time-limited signed
authentication parameter which lets nginx fetch the non-public file.
This adds a number of nginx-level configuration parameters to control
the caching which nginx performs, including the amount of in-memory
index for he cache, the maximum storage of the cache on disk, and how
long data is retained in the cache. The currently-chosen figures are
reasonable for small to medium deployments.
The most notable effect of this change is in allowing browsers to
cache uploaded image content; however, while there will be many fewer
requests, it also has an improvement on request latency. The
following tests were done with a non-AWS client in SFO, a server and
S3 storage in us-east-1, and with 100 requests after 10 requests of
warm-up (to fill the nginx cache). The mean and standard deviation
are shown.
| | Redirect to S3 | Caching proxy, hot | Caching proxy, cold |
| ----------------- | ------------------- | ------------------- | ------------------- |
| Time in Django | 263.0 ms ± 28.3 ms | 258.0 ms ± 12.3 ms | 258.0 ms ± 12.3 ms |
| Small file (842b) | 586.1 ms ± 21.1 ms | 266.1 ms ± 67.4 ms | 288.6 ms ± 17.7 ms |
| Large file (660k) | 959.6 ms ± 137.9 ms | 609.5 ms ± 13.0 ms | 648.1 ms ± 43.2 ms |
The hot-cache performance is faster for both large and small files,
since it saves the client the time having to make a second request to
a separate host. This performance improvement remains at least 100ms
even if the client is on the same coast as the server.
Cold nginx caches are only slightly slower than hot caches, because
VPC access to S3 endpoints is extremely fast (assuming it is in the
same region as the host), and nginx can pool connections to S3 and
reuse them.
However, all of the 648ms taken to serve a cold-cache large file is
occupied in nginx, as opposed to the only 263ms which was spent in
nginx when using redirects to S3. This means that to overall spend
less time responding to uploaded-file requests in nginx, clients will
need to find files in their local cache, and skip making an
uploaded-file request, at least 60% of the time. Modeling shows a
reduction in the number of client requests by about 70% - 80%.
The `Content-Disposition` header logic can now also be entirely shared
with the local-file codepath, as can the `url_only` path used by
mobile clients. While we could provide the direct-to-S3 temporary
signed URL to mobile clients, we choose to provide the
served-from-Zulip signed URL, to better control caching headers on it,
and greater consistency. In doing so, we adjust the salt used for the
URL; since these URLs are only valid for 60s, the effect of this salt
change is minimal.
Moving `/user_avatars/` to being served partially through Django
removes the need for the `no_serve_uploads` nginx reconfiguring when
switching between S3 and local backends. This is important because a
subsequent commit will move S3 attachments to being served through
nginx, which would make `no_serve_uploads` entirely nonsensical of a
name.
Serve the files through Django, with an offload for the actual image
response to an internal nginx route. In development, serve the files
directly in Django.
We do _not_ mark the contents as immutable for caching purposes, since
the path for avatar images is hashed only by their user-id and a salt,
and as such are reused when a user's avatar is updated.
Importing `upload_backend` directly means that in testing it must also
be mocked where it is imported, in order to correctly test the right
backend. Since `get_avatar_url` is part of the public
`ZulipUploadBackend` API, add another helper method to call that.
The `django-sendfile2` module unfortunately only supports a single
`SENDFILE` root path -- an invariant which subsequent commits need to
break. Especially as Zulip only runs with a single webserver, and
thus sendfile backend, the functionality is simple to inline.
It is worth noting that the following headers from the initial Django
response are _preserved_, if present, and sent unmodified to the
client; all other headers are overridden by those supplied by the
internal redirect[^1]:
- Content-Type
- Content-Disposition
- Accept-Ranges
- Set-Cookie
- Cache-Control
- Expires
As such, we explicitly unset the Content-type header to allow nginx to
set it from the static file, but set Content-Disposition and
Cache-Control as we want them to be.
[^1]: https://www.nginx.com/resources/wiki/start/topics/examples/xsendfile/
Enforcing a consistent `type` helps us double-check that we're not
playing fast-and-loose with any file paths for local files. As noted
in the comment, this is purely for defense-in-depth.
Passing `write_local_file` a consistent `type` requires removing the
"avatars" out of `realm_avatar_and_logo_path` -- which makes it
consistent across upload backends.
This, in turn, requires a compensatory change to zerver.lib.export, to
be explicit that the realm icons are exported from the avatars
directory. This clarity is likely an improvement.
sendfile already applied a Content-Disposition header, but the
algorithm may provide both `filename=` and `filename*=` values (which
is potentially confusing to clients) and incorrectly slash-escapes
quotes in Unicode strings.
Django provides a correct implementation, but it is only accessible to
FileResponse objects. Since the entire point is to offload the
filehandle handling, we cannot use a FileResponse.
Django 4.2 will make the function available outside of FileResponse.
Until then, extract our own Content-Disposition handling, based on
Django's.
We remove the very verbose comment added in d4360e2287, describing
Content-Disposition headers, as it does not add much.
As uploads are a feature of the application, not of a generic nginx
deployment, move them into the `zulip::app_frontend_base` class. This
is purely for organizational clarity -- we do not support deployments
with has `zulip::nginx` but not `zulip::app_frontend_base`.
Previously, `templates/zerver/api/missing.md` was a symlink to the
file in the help center documentation.
Prep commit for moving the help center documentation to a top level
directory.
Add more tests analogous to existing ones but for different scenarios.
This is mostly boring text, but is important for completeness, since the
notificability logic underneath is subtle.
Split the one giant `test_end_to_end_missedmessage_hook` into many
smaller tests.
This allows us to not worry about resetting database state after each
test case and also allows extracting a lot of common stuff into setUp
and tearDown.
There is probably even more scope of deduplication here (for example,
the mock and the `assert_maybe_enqueue_notifications_call_args` call are
same for all test cases) but that might not be worth the added
complexity.
We also change a few
```
user_profile.<setting> = <value>
user_profile.save()
```
expressions to instead use the `do_change_user_setting` function.
For alert words, we currently don't send email/push notifications --
only desktop notifications. Thus, we don't need to consider alert words
here, since desktop notifications do not utilize the presence status
calculated at this stage.
Tested manually that alert word desktop notifications work as expected.
When we implement email/push notifications for alert words (issues #5137
and #13127), we can add new fields like
`notifications_data.alert_word_email_notify`, similar to the existing
`notifications_data.wildcard_mention_email_notify`, which will allow us
to keep the alert word notifiability check inside the dataclass, similar
to how the mentions checks are done currently. So, even when that
feature is implemented, the code which this commit removes would be
unnecessary.
We now use the undo supporting `insert` function from the
`text-field-edit` module to update the text after autocompletion,
instead of just resetting the value of the textarea / input, which was
not undo-able.
So that the undo history seems sensible, we replace only the minimal
diff between the text before and after autocompletion. This ensures
that only this diff, and not the entire text, is highlighted when
undoing, as would be ideal.
Fixes: #14311.
Sahil Batra
Lauryn Menard
Aman Agrawal
Josh Klar
Manas Jayaswal
Tim Abbott
Alex Vandiver
Mateusz Mandera
Abhijeet Prasad Bodas
N-Shar-ma