This commit removes the 'experimental_preconnect_link_header' site setting, and the 'preload_link_header' site setting, and introduces two new global settings: early_hint_header_mode and early_hint_header_name.
We don't actually send 103 Early Hint responses from Discourse. However, upstream proxies can be configured to cache a response header from the app and use that to send an Early Hint response to future clients.
- `early_hint_header_mode` specifies the mode for the early hint header. Can be nil (disabled), "preconnect" (lists just CDN domains) or "preload" (lists all assets).
- `early_hint_header_name` specifies which header name to use for the early hint. Defaults to "Link", but can be changed to support different proxy mechanisms.
Why this change?
According to https://web.dev/articles/preload-critical-assets,
> By preloading a certain resource, you are telling the browser that you would like to fetch it sooner than the browser would otherwise discover it because you are certain that it is important for the current page.
The preload resource hint is meant to tell the browser to fetch
resources that it would not discover upfront or early. However, we are
not using it the right way because we are literally adding the resource
hint right before a `<script>` tag which means the browser would have
discovered the resource even without the resource hint.
What does this change do?
This commit removes the preload resource hint which are added right
before script tags since the optimization here is highly questionable at the expense of making
our initial DOM larger.
Sometimes we add scripts outside of Rails. This commit provides a way to generate a nonce placeholder even if you don't have access to an ApplicationController instance.
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.
Safari has a bug which means that scripts with the `defer` attribute are executed before stylesheets have finished loading. This is being tracked at https://bugs.webkit.org/show_bug.cgi?id=209261.
This commit works around the problem by introducing a no-op inline `<script>` to the end of our HTML document. This works because defer scripts are guaranteed to run after inline scripts, and inline scripts are guaranteed to run after any preceding stylesheets.
Technically we only need this for Safari. But given that the cost is so low, it makes sense to include it everywhere rather than incurring the complexity of gating it by user-agent.
Previously, the app HTML served by the Ember-CLI proxy was generated based on a 'bootstrap json' payload generated by Rails. This inevitably leads to differences between the Rails HTML and the Ember-CLI HTML.
This commit overhauls our proxying strategy. Now, we totally ignore the ember-cli `index.html` file. Instead, we take the full HTML from Rails and surgically replace script URLs based on a `data-discourse-entrypoint` attribute. This should be faster (only one request to Rails), more robust, and less confusing for developers.
With Embroider, we can rely on async `import()` to do the splitting
for us.
This commit extracts from `pretty-text` all the parts that are
meant to be loaded async into a new `discourse-markdown-it` package
that is also a V2 addon (meaning that all files are presumed unused
until they are imported, aka "static").
Mostly I tried to keep the very discourse specific stuff (accessing
site settings and loading plugin features) inside discourse proper,
while the new package aims to have some resembalance of a general
purpose library, a MarkdownIt++ if you will. It is far from perfect
because of how all the "options" stuff work but I think it's a good
start for more refactorings (clearing up the interfaces) to happen
later.
With this, pretty-text and app/lib/text are mostly a kitchen sink
of loosely related text processing utilities.
After the refactor, a lot more code related to setting up the
engine are now loaded lazily, which should be a pretty nice win. I
also noticed that we are currently pulling in the `xss` library at
initial load to power the "sanitize" stuff, but I suspect with a
similar refactoring effort those usages can be removed too. (See
also #23790).
This PR does not attempt to fix the sanitize issue, but I think it
sets things up on the right trajectory for that to happen later.
Co-authored-by: David Taylor <david@taylorhq.com>
When Discourse first introduced brotli support, reverse-proxy/CDN support for passing through the accept-encoding header to our NGINX server was very poor. Therefore, a separate `/brotli_assets/...` path was introduced to serve the brotli assets. This worked well, but introduces additional complexity and inconsistencies.
Nowadays, Brotli encoding is well supported, so we don't need the separate paths any more. Requests can be routed to the asset `.js` URLs, and NGINX will serve the brotli/gzip version of the asset automatically.
These have been deprecated for some time, and the vast majority of themes/plugins have already removed their use. The prototype extensions were unexpectedly disabled as a side effect of 895036bd7a (more details in https://github.com/discourse/discourse/pull/24101).
Given that restoring the functionality now involves significant complexity, and would only be delaying the inevitable removal in a matter of months, we've decided to keep them disabled. This commit explicitly sets the flag in the ember environment config to make things clearer.
Previously we were patching ember-cli so that it would split the test bundle into two halves: the helpers, and the tests themselves. This was done so that we could use the helpers for `/theme-qunit` without needing to load all the core tests. This patch has proven problematic to maintain, and will become even harder under Embroider.
This commit removes the patch, so that ember-cli goes back to generating a single `tests.js` bundle. This means that core test definitions will now be included in the bundle when using `/theme-qunit`, and so this commit also updates our test module filter to exclude them from the run. This is the same way that we handle plugin tests on the regular `/tests` route, and is fully supported by qunit.
For now, this keeps `/theme-qunit` working in both development and production environments. However, we are very likely to drop support in production as part of the move to Embroider.
Our Ember build compiles assets into multiple chunks. In the past, we used the output from ember-auto-import-chunks-json-generator to give Rails a map of those chunks. However, that addon is specific to ember-auto-import, and is not compatible with Embroider.
Instead, we can switch to parsing the html files which are output by ember-cli. These are guaranteed to have the correct JS files in the correct place. A <discourse-chunked-script> will allow us to easily identify which chunks belong to which entrypoint.
In future, as we update more entrypoints to be compiled by Embroider/Webpack, we can easily introduce new wrappers.
Previously applied in 2c58d45 and reverted in 24d46fd. This version has been updated for subfolder support.
Our Ember build compiles assets into multiple chunks. In the past, we used the output from `ember-auto-import-chunks-json-generator` to give Rails a map of those chunks. However, that addon is specific to ember-auto-import, and is not compatible with Embroider.
Instead, we can switch to parsing the html files which are output by ember-cli. These are guaranteed to have the correct JS files in the correct place. A `<discourse-chunked-script>` will allow us to easily identify which chunks belong to which entrypoint.
In future, as we update more entrypoints to be compiled by Embroider/Webpack, we can easily introduce new wrappers.
In e1d27400f5 we started running the splash-screen JS through terser, which removed the trailing newline from the `sourceMappingURL` line.
Adding a reliable end-to-end test for this isn't possible because our testing environment doesn't use terser.
The #pluck_first freedom patch, first introduced by @danielwaterworth has served us well, and is used widely throughout both core and plugins. It seems to have been a common enough use case that Rails 6 introduced it's own method #pick with the exact same implementation. This allows us to retire the freedom patch and switch over to the built-in ActiveRecord method.
There is no replacement for #pluck_first!, but a quick search shows we are using this in a very limited capacity, and in some cases incorrectly (by assuming a nil return rather than an exception), which can quite easily be replaced with #pick plus some extra handling.
Currently when generating a onebox for Discourse topics, some important
context is missing such as categories and tags.
This patch addresses this issue by introducing a new onebox engine
dedicated to display this information when available. Indeed to get this
new information, categories and tags are exposed in the topic metadata
as opengraph tags.
If configured, this will be used for static JS assets which are stored on S3. This can be useful if you want to use different CDN providers/configuration for Uploads and JS
Meta topic: https://meta.discourse.org/t/meta-theme-color-is-not-respecting-current-color-scheme/239815
Currently, the dark mode theme-color `<meta>` tag doesn't apply because the light mode tag has `media="all"`. This means that the dark mode `<meta>` tag with `media="(prefers-color-scheme: dark)"` won't override it. This PR updates the light mode tag to `media="(prefers-color-scheme: light)"` if `dark_scheme_id` is defined and leaves it as `media="all"` otherwise.
Meta topic: https://meta.discourse.org/t/meta-theme-color-is-not-respecting-current-color-scheme/239815/7?u=osama.
This commit renders an additional `theme-color` `<meta>` tag for the dark scheme if the current user/request has a scheme selected for dark mode. We currently only render one `theme-color` tag which is always based on the user's selected scheme for light mode, but if the user also selects a scheme for dark mode and uses a device that's configured to use/prefer dark mode, the Discourse UI will be in dark mode, but any parts of the browser/OS UI that's colored based on the `theme-color` tag, would use a color from the user's selected light scheme and look inconsistent with the Discourse UI because the `theme-color` tag is based on the user's selected light scheme.
The additional `theme-color` tag has `media="(prefers-color-scheme: dark)"` and is based on the user's selected dark scheme which means any browser UI that's colored based on `theme-color` tags should be able to pick the right tag based on the user's preference for light/dark mode.
* Revert "Revert "FEATURE: Preload resources via link header (#18475)" (#18511)"
This reverts commit 95a57f7e0c.
* put behind feature flag
* env -> global setting
* declare global setting
* forgot one spot
Experiment moving from preload tags in the document head to preload information the the response headers.
While this is a minor improvement in most browsers (headers are parsed before the response body), this allows smart proxies like Cloudflare to "learn" from those headers and build HTTP 103 Early Hints for subsequent requests to the same URI, which will allow the user agent to download and parse our JS/CSS while we are waiting for the server to generate and stream the HTML response.
Co-authored-by: Penar Musaraj <pmusaraj@gmail.com>
This commit renames all secure_media related settings to secure_uploads_* along with the associated functionality.
This is being done because "media" does not really cover it, we aren't just doing this for images and videos etc. but for all uploads in the site.
Additionally, in future we want to secure more types of uploads, and enable a kind of "mixed mode" where some uploads are secure and some are not, so keeping media in the name is just confusing.
This also keeps compatibility with the `secure-media-uploads` path, and changes new
secure URLs to be `secure-uploads`.
Deprecated settings:
* secure_media -> secure_uploads
* secure_media_allow_embed_images_in_emails -> secure_uploads_allow_embed_images_in_emails
* secure_media_max_email_embed_image_size_kb -> secure_uploads_max_email_embed_image_size_kb
This lets us use all our normal JS tooling like prettier, esline and babel on the splash screen JS. At runtime the JS file is read and inlined into the HTML. This commit also switches us to use a CSP hash rather than a nonce for the splash screen.
- `no_custom` -> `no_themes` (history: before themes existed, we had a similar tool called 'customizations')
- `only_official` -> `no_unofficial_plugins` (matches format of `no_themes` and `no_plugins`, and makes it clear that this doesn't affect themes)
- `?safe_mode=no_themes%2C%no_plugins` -> `?safe_mode=no_themes,no_plugins` (the query portion of a URL does not require commas to be encoded. This is much nicer to read)
- If `no_plugins` is chosen from `/safe-mode` the URL generated will omit the superfluous `no_unofficial_plugins` flag
- Some tweaks to copy on `/safe-mode`
* FIX: bots could generate errors when slug generation method is encoded
When slug generation method is encoded (non default) then bots could
cause errors in the logs for urls containing special chars.
ó for example in a URL can be requested in a valid ASCII-8BIT string, and
later when joined to UTF-8 would result in encoding issues.
Fix here ensures we force encoding correctly for outlier cases.
Browser tend to always encode these chars, hence we did not notice this.
Co-authored-by: Jarek Radosz <jradosz@gmail.com>
All our development-mode assets serve a `Cache-Control: no-cache` header, so a query parameter shouldn't be needed. Ember CLI does not include cache-busting parameters, so this change will move the development rails app to the same behaviour.
This will fix adding persistent breakpoints in the dev tools. Previously, the browser would think that the assets have been replaced and throw away the breakpoints.
This commit adds preload links for core/plugin/theme CSS stylesheets in the head.
Preload links are non-blocking and run in parallel. This means that they should have already been downloaded by the time we use the actual stylesheets (in the <body> tag).
Google is currently complaining about this here and this PR will address that warning.
This commit will also fix an issue in the splash screen where it sometimes doesn't respect the theme colors - causing a slightly jarring experience on dark themes.
Note that I opted not to add new specs because the underlying work required already has a lot of coverage. The new methods only change the output HTML so we can chuck that in the document <head>
This change also means that we can make all the stylesheets non-render blocking, but that will follow in a separate commit.
We previously relied on CSS animation-delay for the splash. This means that we can get inconsistent results based on device/network conditions.
This PR moves us to a more consistent timing based on {request time + 2 seconds}
Internal topic: /t/65378/65
This PR introduces a new hidden site setting that allows admins to display a splash screen while site assets load.
The splash screen can be enabled via the `splash_screen` hidden site setting.
This is what the splash screen currently looks like
5ceb72f085.mp4
Once site assets load, the splash screen is automatically removed.
To control the loading text that shows in the splash screen, you can change the preloader_text translation string in admin > customize > text
This is pre-request work to introduce a splash screen while site assets load.
The only change this commit introduces is that it ensures we add the defer attribute to core/plugin/theme .JS files. This will allow us to insert markup before the browser starts evaluating those scripts later on. It has no visual or functional impact on core.
This will not have any impact on how themes and plugins work. The only exception is themes loading external scripts in the </head> theme field directly via script tags. Everything will work the same but those would need to add the defer attribute if they want to keep the benefits introduced in this PR.
Twitter does not allow SVGs to be used for twitter:image
metadata (see https://developer.twitter.com/en/docs/twitter-for-websites/cards/overview/markup)
so we should fall back to the site logo if the image option
provided to `crawlable_meta_data` or SiteSetting.site_twitter_summary_large_image_url
is an SVG, and do not add the meta tag for twitter:image at all
if the site logo is an SVG.
