When "lazy load categories" is enabled, only the categories present in
the sidebar are preloaded. This is insufficient because the parent
categories are necessary too for the sidebar to be rendered properly.
The strict-dynamic CSP directive is supported in all our target browsers, and makes for a much simpler configuration. Instead of allowlisting paths, we use a per-request nonce to authorize `<script>` tags, and then those scripts are allowed to load additional scripts (or add additional inline scripts) without restriction.
This becomes especially useful when admins want to add external scripts like Google Tag Manager, or advertising scripts, which then go on to load a ton of other scripts.
All script tags introduced via themes will automatically have the nonce attribute applied, so it should be zero-effort for theme developers. Plugins *may* need some changes if they are inserting their own script tags.
This commit introduces a strict-dynamic-based CSP behind an experimental `content_security_policy_strict_dynamic` site setting.
Followup to 978d52841a
It's complicated...we have multiple "anonymous" user concepts
in core, and even two classes called the exact same thing --
AnonymousUser.
The first case is Guardian::AnonymousUser, which is used for
people who are browsing the forum without being authenticated.
The second case is the model AnonymousUser, which is used when
a user is liking or posting anonymously via allow_anonymous_likes
or allow_anonymous_posting site settings.
We will untangle this naming nightmare later on...but for the
time being, only authenticated users who are pretending to be
anonymous should be able to like posts if allow_anonymous_likes
is on.
Why this change?
This is a first pass at adding an objects validator which main's job is
to validate an object against a defined schema which we will support. In
this pass, we are simply validating that properties that has been marked
as required are present in the object.
Reactions needs this to be able to filter out likes received
actions, where there is also an associated reaction, since
now most reactions also count as a like.
Why this change?
This is a follow up to c30aeafd9d. The
commit was calling `BlockRequestsMiddleware.allow_requests!` only before
`type: :system` tests but non system type tests could be running as well
and needs the `BlockRequestsMiddleware.allow_requests!` middleware to be
disabled too.
When making sensitive changes to an account (adding 2FA or passkeys), we
require users to confirm their password. This is to prevent an attacker
from adding 2FA to an account they have access to.
However, on newly created accounts, we should not require this, it's an
extra step and it doesn't provide extra security (since the account was
just created). This commit makes it so that we don't require session
confirmation for accounts created less than 5 minutes ago.
Why this change?
We have been debugging flaky system tests and noticed in https://github.com/discourse/discourse/actions/runs/7911902047/job/21596791343?pr=25690
that ActiveRecord connection checkout timeouts are encountered because
the Capybara server thread is processing requests even after
`Capybara.reset_session!` and ActiveRecord's `teardown_fixtures` have already been call.
The theory here is that an inflight request can still hit the Capybara
server even after `Capybara.reset_session!` has been called and end up
eating up an ActiveRecord connection for too long and also messing with
the database outside of a transaction.
What does this change do?
This change adds a `BlockRequestsMiddleware` middleware in the test
environment which is enabled to reject all incoming requests at the end
of each system test and before `Capybara.reset_session!` is called. At
the start of each RSpec test, the middleware is disabled again.
Why this change?
On CI, we have been seeing flaky system tests because ActiveRecord is
unable to checkout a connection. This patch is meant to help us debug
which thread is not returning the connection to the queue.
This would allow a theme component (or an API call) to reset the bump
date of a topic to a given post's created_at date.
I picked `post_id` as the parameter here because it provides a bit of
extra protection against accidentally resetting the bump date to a date
that doesn't make sense.
Fixes an issue where private topics that are quoted have an incorrectly formatted url when using a subfolder install.
This update returns a relative url that includes the base_path rather than a combination of base_url + base_path.
This commit includes several changes to make hashtags work when "lazy
load categories" is enabled. The previous hashtag implementation use the
category colors CSS variables, but these are not defined when the site
setting is enabled because categories are no longer preloaded.
This commit implements two fundamental changes:
1. load colors together with the other hashtag information
2. load cooked hashtag data asynchronously
The first change is implemented by adding "colors" to the HashtagItem
model. It is a list because two colors are returned for subcategories:
the color of the parent category and subcategory.
The second change is implemented on the server-side in a new route
/hashtags/by-ids and on the client side by loading previously unseen
hashtags, generating the CSS on the fly and injecting it into the page.
There have been minimal changes outside of these two fundamental ones,
but a refactoring will be coming soon to reuse as much of the code
and maybe favor use of `style` rather than injecting CSS into the page,
which can lead to page rerenders and indefinite grow of the styles.
This commit changes `max_image_megapixels` to be used
as is without multiplying by 2 to give extra leway.
We found in reality this was just causing confusion
for admins, especially with the already permissive
40MP default.
When we show the links to installed plugins in the admin
sidebar (for plugins that have custom admin routes) we were
previously only doing this if you opened /admin, not if you
navigated there from the main forum. We should just always
preload this data if the user is admin.
This commit also changes `admin_sidebar_enabled_groups` to
not be sent to the client as part of ongoing efforts to
not check groups on the client, since not all a user's groups
may be serialized.
When we insert into the hot set we add things with a score of 0
This means that if hot has more than batch size items in it with a score, then the 0s don't get an initial score
This corrects the situation by always ensuring we re-score:
1. batch size high scoring topics
2. (new) batch size recently bumped topics
* Update spec/models/topic_hot_scores_spec.rb
Co-authored-by: Isaac Janzen <50783505+janzenisaac@users.noreply.github.com>
---------
Co-authored-by: Isaac Janzen <50783505+janzenisaac@users.noreply.github.com>
This reverts commit 767b49232e.
If anything else (e.g. GTM integration) introduces a nonce/hash, then this change stops the splash screen JS to fail and makes sites unusable.
Removes duplication from LimitedEdit to see who can edit
posts, and also removes the old trust level setting check
since it's no longer necessary.
Also make it so staff can always edit since can_edit_post?
already has a staff escape hatch.
Why this change?
This commit introduces an experimental `type: objects` theme setting
which will allow theme developers to store a collection of objects as
JSON in the database. Currently, the feature is still in development and
this commit is simply setting up the ground work for us to introduce the
feature in smaller pieces.
What does this change do?
1. Adds a `json_value` column as `jsonb` data type to the `theme_settings` table.
2. Adds a `experimental_objects_type_for_theme_settings` site setting to
determine whether `ThemeSetting` records of with the `objects` data
type can be created.
3. Updates `ThemeSettingsManager` to support read/write access from the
`ThemeSettings#json_value` column.
Followup fb087b7ff6
post_links_allowed_groups is an odd check tied to
unrestricted_link_posting? in PostGuardian, in that
it doesn't have an escape hatch for staff like most
of the rest of these group based settings.
It doesn't make sense to exclude admins or mods from
posting links, so just always allow them to avoid confusion.
Browsers will ignore unsafe-inline if nonces or hashes are included in the CSP. When unsafe-inline is enabled, nonces and hashes are not required, so we can skip them.
Our strong recommendation remains that unsafe-inline should not be used in production.
Why this change?
This is caused by a regression in
59839e428f, where we stopped saving the
`Theme` object because it was unnecessary. However, it resulted in the
`after_save` callback not being called and hence
`Theme#update_javascript_cache!` not being called. As a result, some
sites were reporting that after runing a theme migration, the defaults
for the theme settings were used instead of the settings overrides
stored in the database.
What does this change do?
Add a call to `Theme#update_javascript_cache!` after running theme
migrations.
This fixes a bug where the sidebar categories would not be loaded when
the categories were lazy loaded because the sidebar uses the preloaded
category list, which was empty.
We just completed the 3.2 release, which marks a good time to drop some previously deprecated columns.
Since the column has been marked in ignored_columns, it has been inaccessible to application code since then. There's a tiny risk that this might break a Data Explorer query, but given the nature of the column, the years of disuse, and the fact that such a breakage wouldn't be critical, we accept it.
Why this change?
We are getting the following error on CI:
`Text file busy -
/github/home/.cache/selenium/chromedriver/linux64/121.0.6167.85/chromedriver`
This happens when two processes tries to download the chromedriver at
the same time. I'm not entirely sure why the previous implementatio is
not locking properly since we still saw the `Text file busy` error but
by accquring the lock before we even check for the existence of
`~/.cache/selenium`, we should be able to eliminate the chance of two
processes trying to download the chromedriver binary at the same time.
1. Don't show visited line for hot filter, it is in random order
2. Don't count likes on non regular posts (eg: whispers / small actions)
3. Don't count participants in non regular posts
Checking group permissions on the client does not work,
since not all groups are serialized to the client all
the time. We can check `uploaded_avatars_allowed_groups`
on the server side and serialize to the current user
instead.
CI runs on slower machines, so we need to use longer wait times. `Capybara.default_max_wait_time` is automatically reconfigured based on the environment.
JS assets added by plugins via `register_asset` will be outside the `assets/javascripts` directory, and are therefore exempt from being transpiled. That means that there isn't really any need to run them through DiscourseJsProcessor. Instead, we can just concatenate them together, and avoid the need for all the sprockets-wrangling.
This commit also takes the opportunity to clean up a number of plugin-asset-related codepaths which are no longer required (e.g. globs, handlebars)
1. Serial likers will just like a bunch of posts on the same topic, this will
heavily inflate hot score. To avoid artificial "heat" generated by one user only count
the first like on the topic within the recent_cutoff range per topic
2. When looking at recent topics prefer "unique likers", defer to total likes on
older topics cause we do not have an easy count for unique likers
3. Stop taking 1 off like_count, it is not needed - platforms like reddit
allow you to like own post so they need to remove it.
Why this change?
Returning an array makes it hard to immediately retrieve a setting by
name and makes the retrieval an O(N) operation. By returning an array,
we make it easier for us to lookup a setting by name and retrieval is
O(1) as well.
Why this change?
Since 1dba1aca27, we have been remapping
the `<->` proximity operator in a tsquery to `&`. However, there is
another variant of it which follows the `<N>` pattern. For example, the
following text "end-to-end" will eventually result in the following
tsquery `end-to-end:* <-> end:* <2> end:*` being generated by Postgres.
Before this fix, the tsquery is remapped to `end-to-end:* & end:* <2>
end:*` by us. This is requires the search data which we store to contain
`end` at exactly 2 position apart. Due to the way we limit the
number of duplicates in our search data, the search term may end up not
matching anything. In bd32912c5e, we made
it such that we do not allow any duplicates when indexing a topic's
title. Therefore, search for `end-to-end` against a topic title with
`end-to-end` will never match because our index will only contain one
`end` term.
What does this change do?
We will remap the `<N>` variant of the proximity operator.
