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.
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.
Currently, `window.I18n` is defined in an old school hand written
script, inlined into locale/*.js by the Rails asset pipeline, and
then the global variable is shimmed into a pseudo AMD module later
in `module-shims.js`.
This approach has some problems – for one thing, when we add a new
V2 addon (e.g. in #23859), Embroider/Webpack is stricter about its
dependencies and won't let you `import from "I18n";` when `"I18n"`
isn't listed as one of its `dependencies` or `peerDependencies`.
This moves `I18n` into a real package – `discourse-i18n`. (I was
originally planning to keep the `I18n` name since it's a private
package anyway, but NPM packages are supposed to have lower case
names and that may cause problems with other tools.)
This package defines and exports a regular class, but also defines
the default global instance for backwards compatibility. We should
use the exported class in tests to make one-off instances without
mutating the global instance and having to clean it up after the
test run. However, I did not attempt that refactor in this PR.
Since `discourse-i18n` is now included by the app, the locale
scripts needs to be loaded after the app chunks. Since no "real"
work happens until later on when we kick things off in the boot
script, the order in which the script tags appear shouldn't be a
problem. Alternatively, we can rework the locale bundles to be more
lazy like everything else, and require/import them into the app.
I avoided renaming the imports in this commit since that would be
quite noisy and drowns out the actual changes here. Instead, I used
a Webpack alias to redirect the current `"I18n"` import to the new
package for the time being. In a separate commit later on, I'll
rename all the imports in oneshot and remove the alias. As always,
plugins and the legacy bundles (admin/wizard) still relies on the
runtime AMD shims regardless.
For the most part, I avoided refactoring the actual I18n code too
much other than making it a class, and some light stuff like `var`
into `let`.
However, now that it is in a reasonable format to work with (no
longer inside the global script context!) it may also be a good
opportunity to refactor and make clear what is intended to be
public API vs internal implementation details.
Speaking of, I took the librety to make `PLACEHOLDER`, `SEPARATOR`
and `I18nMissingInterpolationArgument` actual constants since it
seemed pretty clear to me those were just previously stashed on to
the `I18n` global to avoid polluting the global namespace, rather
than something we expect the consumers to set/replace.
* 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 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.
This commit does six things
* changes the animation for the splash screen. To a more subtle animation.
* defers displaying the splash by 1.5 seconds
* defers displaying the splash "loading" text by 2.5 seconds
* defers removing the splash until all Discourse initializers have run
* fixes a display issue in Firefox
* Inlines the SVG as a base64 and inlines the required CSS.
The encoded SVG is hard coded for now, but we will use a helper to generate that based on the file after some testing.
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
Similar to #17145
This commit moves the SVG sprite container to the <discourse-assets> element.
There is 0 visual or functional changes in this PR. It just tidies up the element view in devTools.
This PR introduces 0 visual or functional changes. The only thing that it changes is that it moves the data-preloaded div (which has the app boot json into the <discourse-assets> element.
See #17078 for a bit more context.
The reason behind this change is that it makes devTools element view a little bit less cluttered.
This is related to #17063 and is also a pre-request for the splash screen work.
This PR introduces 0 visual or functional changes. It just relocates the stylesheets in the load order.
`.css` stylesheets block the browser render. We need to move those out of the <head> tag.
However, they still need to be loaded before core/plugin/theme rendered HTML to avoid FOUC.
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.
Previously, accessing the Rails app directly in development mode would give you assets from our 'legacy' Ember asset pipeline. The only way to run with Ember CLI assets was to run ember-cli as a proxy. This was quite limiting when working on things which are bypassed when using the ember-cli proxy (e.g. changes to `application.html.erb`). Also, since `ember-auto-import` introduced chunking, visiting `/theme-qunit` under Ember CLI was failing to include all necessary chunks.
This commit teaches Sprockets about our Ember CLI assets so that they can be used in development mode, and are automatically collected up under `/public/assets` during `assets:precompile`. As a bonus, this allows us to remove all the custom manifest modification from `assets:precompile`.
The key changes are:
- Introduce a shared `EmberCli.enabled?` helper
- When ember-cli is enabled, add ember-cli `/dist/assets` as the top-priority Rails asset directory
- Have ember-cli output a `chunks.json` manifest, and teach `preload_script` to read it and append the correct chunks to their associated `afterFile`
- Remove most custom ember-cli logic from the `assets:precompile` step. Instead, rely on Rails to take care of pulling the 'precompiled' assets into the `public/assets` directory. Move the 'renaming' logic to runtime, so it can be used in development mode as well.
- Remove fingerprinting from `ember-cli-build`, and allow Rails to take care of things
Long-term, we may want to replace Sprockets with the lighter-weight Propshaft. The changes made in this commit have been made with that long-term goal in mind.
tldr: when you visit the rails app directly, you'll now be served the current ember-cli assets. To keep these up-to-date make sure either `ember serve`, or `ember build --watch` is running. If you really want to load the old non-ember-cli assets, then you should start the server with `EMBER_CLI_PROD_ASSETS=0`. (the legacy asset pipeline will be removed very soon)
These were originally very similar, but have diverged over time. This makes it very difficult to manage styling.
This commit moves the noscript header and footer into partials so they can be reused in both the crawler view and the `<noscript>` view. It also makes browser-update render the noscript content **instead of** the `<section id='main'>`, rather than adding adding the noscript inside the `<section>`. This provides better parity with the server-rendered crawler view.
This makes a small improvement to 'cold cache' ember-cli build times, and a large improvement to 'warm cache' build times
The ember-auto-import update means that vendor is now split into multiple files for efficiency. These are named `chunk.*`, and should be included immediately after the `vendor.js` file. This commit also updates the rails app to render script tags for these chunks.
This change was previously merged, and caused memory-related errors on RAM-constrained machines. This was because Webpack 5 switches from multiple worker processes to a single multi-threaded process. This meant that it was hitting node's default heap size limit (~500mb on a 1GB RAM server). Discourse's standard install procedure recommends adding 2GB swap to 1GB-RAM machines, so we can afford to override's Node's default via the `--max-old-space-size` flag.
This reverts commit f4c6a61855 and a8325c9016
This update of ember-auto-import and webpack causes significantly higher memory use during rebuilds. This made ember-cli totally unusable on 1GB RAM / 2GB swap environments. We don't have a specific need for this upgrade right now, so reverting for now.
This makes a small improvement to 'cold cache' ember-cli build times, and a large improvement to 'warm cache' build times
The ember-auto-import update means that vendor is now split into multiple files for efficiency. These are named `chunk.*`, and should be included immediately after the `vendor.js` file. This commit also updates the rails app to render script tags for these chunks
Under some conditions, these varied responses could lead to cache poisoning, hence the 'security' label.
Previously the Rails application would serve JSON data in place of HTML whenever Ember CLI requested an `application.html.erb`-rendered page. This commit removes that logic, and instead parses the HTML out of the standard response. This means that Rails doesn't need to customize its response for Ember CLI.
We have two JS assets which are included in the `<body>` of responses. We were including the `<link rel='preload'` hint alongside the script tag in the body. Instead, we can move the preload hint to the `<head>` so that the browser discovers it earlier, and can start preloading the assets while the body is loading.
This was problematic if something like SCSS file throws an error as the
app would tell Ember CLI to bootstrap as if everything is fine and not
display the error.
The fix is to only hijack the rendering at the end of the template
instead of the beginning.
This adds an optional ENV variable, `EMBER_CLI_PROD_ASSETS`. If truthy,
compiling production assets will be done via Ember CLI and will replace
the assets Rails would otherwise use.
Before this change, calling `StyleSheet::Manager.stylesheet_details`
for the first time resulted in multiple queries to the database. This is
because the code was modelled in a way where each `Theme` was loaded
from the database one at a time.
This PR restructures the code such that it allows us to load all the
theme records in a single query. It also allows us to eager load the
required associations upfront. In order to achieve this, I removed the
support of loading multiple themes per request. It was initially added
to support user selectable theme components but the feature was never
completed and abandoned because it wasn't a feature that we thought was
worth building.
In Ember CLI addons get put into the vendor bundle, as opposed to their
own bundle like we're doing in the Rails app. We never use pretty-text
without our vendor bundle so this should have no difference on
performance.
We need to keep the pretty-text bundle for server side cooking.
In Ember CLI, the vendor bundler includes Ember/jQuery, so this brings
our app closer to that configuration.
We have a couple pages (Reset Password / Confirm New Email) where we need
`ember_jquery` without vendor so the file still exists for those cases.
* DEV: Give a nicer error when `--proxy` argument is missing
* DEV: Improve Ember CLI's bootstrap logic
Instead of having Ember CLI know which URLs to proxy or not, have it try
the URL with a special header `HTTP_X_DISCOURSE_EMBER_CLI`. If present,
and Discourse thinks we should bootstrap the application, it will
instead stop rendering and return a HTTP HEAD with a response header
telling Ember CLI to bootstrap.
In other words, any time Rails would otherwise serve up the HTML for the
Ember app, it stops and says "no, you do it."
* DEV: Support asset filters by path using a new options object
Without this, Ember CLI's bootstrap would not get the assets it wants
because the path it was requesting was different than the browser path.
This adds an optional request header to fix it.
So far this is only used by the styleguide.
This cookie is only used during login. Having it persist after that can
cause some unusual behavior, especially for sites with short session
lengths.
We were already deleting the cookie following a new signup, but not for
existing users.
This commit moves the cookie deletion logic out of the erb template, and
adds logic and tests to ensure it is always deleted consistently.
Co-authored-by: Jarek Radosz <jradosz@gmail.com>
CSS are blocking resources, so keeping them below JS delays
rendering of the page. CSS should be loaded ASAP.
This change speeds up first contentful paint by 0.2s on localhost.
The slower the device, the bigger the difference could be.
DEV: Replace instances of Discourse.base_uri with Discourse.base_path
This is clearer because the base_uri is actually just a path prefix. This continues the work started in 555f467.
* DEV: To be pedantic, there is more than EMBER in there now
* DEV: Use less globals. Have `Discourse` start in an initializer
* DEV: Remove another global
In order to avoid a boatload of attributeBindings, I moved the root
element of the suggested-topics component into the template. Also,
autoformat their hbs files.
Testing info: https://www.scottohara.me/blog/2018/03/03/landmarks.html#using-screen-readers-to-navigate-landmarks
Additionally, flag modals with aria-modal=true to avoid the screenreader
accidentally escaping the modal. There's no need to ever toggle the
attribute to false, because we display:none the modal root when it's
closed.
Additionally correctly handle cookie path for authentication_data
There were two bugs that exposed an interesting case where two discourse
instances hosted across two subfolder installs in the same domain
with oauth may clash and cause strange redirection on first login:
Log in to example.com/forum1. authentication_data cookie is set with path /
On the first redirection, the current authentication_data cookie is not unset.
Log in to example.com/forum2. In this case, the authentication_data cookie
is already set from forum1 - the initial page load will incorrectly redirect
the user to the redirect URL from the already-stored cookie, to /forum1.
This removes this issue by:
* Setting the cookie for the correct path, and not having it on root
* Correctly removing the cookie on first login
Attempt 2, with more test.
Additionally correctly handle cookie path for authentication_data
There were two bugs that exposed an interesting case where two discourse
instances hosted across two subfolder installs in the same domain
with oauth may clash and cause strange redirection on first login:
Log in to example.com/forum1. authentication_data cookie is set with path /
On the first redirection, the current authentication_data cookie is not unset.
Log in to example.com/forum2. In this case, the authentication_data cookie
is already set from forum1 - the initial page load will incorrectly redirect
the user to the redirect URL from the already-stored cookie, to /forum1.
This removes this issue by:
Setting the cookie for the correct path, and not having it on root
Correctly removing the cookie on first login
Additionally correctly handle cookie path for authentication_data
There were two bugs that exposed an interesting case where two discourse
instances hosted across two subfolder installs in the same domain
with oauth may clash and cause strange redirection on first login:
Log in to example.com/forum1. authentication_data cookie is set with path /
On the first redirection, the current authentication_data cookie is not unset.
Log in to example.com/forum2. In this case, the authentication_data cookie
is already set from forum1 - the initial page load will incorrectly redirect
the user to the redirect URL from the already-stored cookie, to /forum1.
This removes this issue by:
* Setting the cookie for the correct path, and not having it on root
* Correctly removing the cookie on first login
