`Upload#url` is more likely and can change from time to time. When it
does changes, we don't want to have to look through multiple tables to
ensure that the URLs are all up to date. Instead, we simply associate
uploads properly to `UserProfile` so that it does not have to replicate
the URLs in the table.
Minor fixes to add Rails 6 support to Discourse, we now will boot
with RAILS_MASTER=1, all specs pass
Only one tiny deprecation left
Largest change was the way ActiveModel:Errors changed interface a
bit but there is a simple backwards compat way of working it
* Fix header size to fit the viewport
Setting the header to border-box means that the padding is factored in when calculating its width. If this is not done, the header width would be 100% of the viewport width + any padding when set on it. That causes some parts of the header to be rendered off-screen and causes horizontal scrolling.
* prevent code in posts from causing horizontal overflow
We are currently not wrapping code in posts in the crawler view. This affects both unformatted code and inline code blocks. This commit forces such code to wrap in order to prevent horizontal overflow which the Google bot complains about since it causes some content to be rendered outside of the viewport for the mobile crawler.
This change automatically resizes icons for various purposes. Admins can now upload `logo` and `logo_small`, and everything else will be auto-generated. Specific icons can still be uploaded separately if required.
## Core
- Adds an SiteIconManager module which manages automatic resizing and fallback
- Icons are looked up in the OptimizedImage table at runtime, and then cached in Redis. If the resized version is missing for some reason, then most icons will fall back to the original files. Some icons (e.g. PWA Manifest) will return `nil` (because an incorrectly sized icon is worse than a missing icon).
- `SiteSetting.site_large_icon_url` will return the optimized version, including any fallback. `SiteSetting.large_icon` continues to return the upload object. This means that (almost) no changes are required in core/plugins to support this new system.
- Icons are resized whenever a relevant site setting is changed, and during post-deploy migrations
## Wizard
- Allows `requiresRefresh` wizard steps to reload data via AJAX instead of a full page reload
- Add placeholders to the **icons** step of the wizard, which automatically update from the "Square Logo"
- Various copy updates to support the changes
- Remove the "upload-time" resizing for `large_icon`. This is no longer required.
## Site Settings UX
- Move logo/icon settings under a new "Branding" tab
- Various copy changes to support the changes
- Adds placeholder support to the `image-uploader` component
- Automatically reloads site settings after saving. This allows setting placeholders to change based on changes to other settings
- Upload site settings will be assigned a placeholder if SiteIconManager `responds_to?` an icon of the same name
## Dashboard Warnings
- Remove PWA icon and PWA title warnings. Both are now handled automatically.
## Bonus
- Updated the sketch logos to use @awesomerobot's new high-res designs
`/new-topic` redirects back to `/latest`, so the ember router considers this to be a 'refresh'. This triggers the `resetParams()` logic, which would cause the transition to abort, and the composer would never open.
This commit fixes the `resetParams()` logic so that it uses the default values, instead of setting everything to `null`. Therefore the transition is no longer aborted, and the composer opens successfully.
Unfortunately I have not been able to reproduce the issue in a QUnit test.
On busy sites, concurrent requests to insert into post_timings can
occur, which was dealt with using Ruby exceptions.
This moves the handling to PostgreSQL which makes it a bit faster,
and prevents a spam of ERROR in the database logs.
If a tag group is set to only be visible to staff, and is restricted
to a category that is visible by everyone, the tags in the group were
being shown on the /tags page. They weren't visible anywhere else.
This commit fixes it so they don't show on the /tags page.
Previously, when existing composer, the `#main-outlet` element padding was set to zero. This inline style would override any CSS set for that element, causing issues with the mobile footer nav.
The fix removes the inline padding style instead of setting it to zero. It also uses integers for the set values, and removes a duplicate style.
The error displayed when logging into suspended accounts via SSO never includes
the suspension reason, unlike non-SSO logins. By re-using the failed_to_login
method when generating the error message for SSO we can ensure the message is
consistent between the SSO and non-SSO paths.
This is for backwards compatibility purposes. Even if `Upload#url` has a
format that we don't recognize, we should still return the upload object
as long as the upload record is present.
* show likes value in crawler view if count is > 0
* remove <hr> since horizontal line is already provided by css - this removes one of 2 horizontal lines in post crawler view
* Fix handling SNS notifications for AWS SES
This fixes detection of email bounce by:
- removing hard requirement for email ID, ID in webhook msg never equals this in email_log
- gets bounce_score from user stats instead of nonexistent field in webhook msg
* Remove empty line
* Prettify access to EmailLog for parsing SNS notification
Co-Authored-By: SystemZ <SystemZ@users.noreply.github.com>
This is a first step of a performance optimisation, more will follow
Previously we did not properly account for previously read topics while
"rushing" marking times on posts.
The new mechanism now avoids "rushing" sending timings to server if all
the posts were read.
Also to alleviate some server load we only "ping" the server with old timings
once a minute (it used to be every 20 seconds)
This commit adds some improvements to native app banners for iOS and Android
- iOS and Android now have separate settings for native app banners
- app banners will now only show for users on TL1 and up
- app ids are now in a hidden site setting to allow sites to switch to their own app, if desired
- iOS only: the site URL is passed to the app arguments
If you turn it on now, default all users to approved since they were
previously. Also support approving a user that doesn't have a reviewable
record (it will be created first.)
This also includes a refactor to move class method calls to
`DiscourseEvent` into an initializer. Otherwise the load order of
classes makes a difference in the test environment and some settings
might be triggered and others not, randomly.
- moves footer nav to the header on iPads
- disables scrolling events for iPads
- removes footer nav entirely on Chrome PWAs
- toggles DiscourseHub iOS app status bar styling (dark/light) when opening/closing lightboxes and modals
Since enabling pinch-to-zoom in iOS (eae22548de), there was an issue with inputs: Safari auto-zooms inputs with font-size under 16px. Now zooming will be disabled while focus is on an input.
This commit also removes a) a lightbox zoom-enabling event (no longer needed) and b) a comment about iOS zoom issues.
* UX: Render wizard previews for high-DPI displays
Sets up a canvas element of twice the required dimensions, scales all coordinates by 2x, then shrinks the display in css.
* Use window.devicePixelRatio to determine scale factor
Theme developers can include any number of scss files within the /scss/ directory of a theme. These can then be imported from the main common/desktop/mobile scss.
The original implementation felt a bit too short, went with
a much simpler animation which sticks around for 5 seconds.
5 seconds seems to be the standard and the same timing gmail use
* FIX: correctly retrieve 'login required' setting value on wizard
FEATURE: extract 'invite only' setting in a separate checkbox control
* Update invite_only checkbox locale on wizard.
Co-Authored-By: techAPJ <arpit@techapj.com>
As of eae2254, zooming is enabled in mobile. Unfortunately, iOS zoomns in on textareas and inputs with font sizes below 16px. This bumps font size for textareas (i.e. composer) but not yet for inputs.
The compress brotli functionality is no longer optional, this has worked
well for years. The name of the ENV var is also confusing cause it does
not have a `DISCOURSE_` prefix which caused issues with the web upgrader
Brotli support is now unconditionally on
In certain situations the detection method fails. For example, it fails
for a topic containing: a long post (~1600px height), gap, small action
(focus on this), hidden reply.
Previously removing a banner raised a null "banner" in the message bus
channel.
Then Ember.Object.create would fail cause it always expects an object in
constructor.
also add a hard limit of 1000 users per job run so we do not clog the
scheduler
destroyer.destroy has a transaction and this can have some serious complications
with the open record set find_each has going
- Plugin developers using OpenID2.0 should migrate to OAuth2 or OIDC. OpenID2.0 APIs will be removed in v2.4.0
- For sites requiring Yahoo login, it can be implemented using the OpenID Connect plugin: https://meta.discourse.org/t/103632
For more information, see https://meta.discourse.org/t/113249
In certain edge cases, the message bus won't send the message to the
user about the updated review count and it can go out of sync.
This patch synchronizes the review count every time:
1. The user visits the "Needs Review" page
2. Every time the user performs an action
restructure query so it avoids ORs
It appears postgres is picking suboptimal indexes if too many ORs exist
despite how trivial the condition is.
This bypasses conditional in the query and evals them upfront.
On meta for my user this made a 10x perf difference.
This boils down to either having `OR u.admin` or not having `OR u.admin` in
the query.
Note, to avoid race conditions we are setting last_unread to 10 minutes ago
if there is nothing unread.
This is safer in case of in progress transactions
we don't want to lose unread for any window of time.
This optimisation avoids large scans joining the topics table with the
topic_users table.
Previously when a user carried a lot of read state we would have to join
the entire read state with the topics table. This operation would slow down
home page and every topic page. The more read state you accumulated the
larger the impact.
The optimisation helps people who clean up unread, however if you carry
unread from years ago it will only have minimal impact.
A new checkbox has been added to the Tags tab of the category settings modal
which is used when some tags and/or tag groups are restricted to the category,
and all other unrestricted tags should also be allowed.
Default is the same as the previous behaviour: only allow the specified set of
tags and tag groups in the category.
Sometimes sidekiq is so fast that it starts jobs before transactions
have comitted. This patch moves the message bus stuff until after things
have comitted.
Such links might be in present in old PMs. For example, a notification of
outstanding flags.
New PMs should receive the correct link but this prevents 404s in the
other case.
"Rejecting" a user in the queue is equivalent to deleting them, which
would then making it impossible to review rejected users. Now we store
information about the user in the payload so if they are deleted things
still display in the Rejected view.
Secondly, if a user is destroyed outside of the review queue, it will
now automatically "Reject" that queue item.
Conversely, if a user is deactivated the reviewable should automatically
be rejected.
Before this fix, if a user was not active they'd still show in the
review queue but without an "Approve" button which was confusing.
Previously every rebake would remove and recreate records in this table
This caused created_at and updated_at to keep changing
Yes, I know the SQL is somewhat complex, but this makes quote extraction
more efficient cause we do everything in 2 round trips.
This also removes some concurrency protection we should no longer need
Some sites have external URLs that don't even match `%/uploads/%' and
some sites surprise me with URLs that contains the default path when it
is a site in a multisite cluster. We can't do anything about those.
User cards triggered in header were incorrectly positioned in Safari desktop.
Using `position()` instead of `offset()` is more consistent, since header is a fixed element in this scenario.
If the post ids keep loading, we might end up in a situations where
we're always loading the same post ids over and over again without
indexing anything new.
Follow up to daeda80ada.
Adds the parallel_tests gem, and redis/postgres configuration for running rspec tests in parallel. To use:
```
rake parallel:rake[db:create]
rake parallel:rake[db:migrate]
rake parallel:spec
```
This brings the test suite from 12m20s to 3m11s on my macOS machine
Handle the case of https://github.com/discourse/DiscoTOC doing this kind of setup:
```
return {
action: "insertDtoc",
icon: "align-left",
label: themePrefix("insert_table_of_contents"),
condition: !composerController.get("model.canCategorize")
};
```
In this case there's no function to call, it's already set.
This commit fixes the follow quality issue with `PostSearchData#raw_data`:
1. URLs are being tokenized and links with similar href and characters
are being duplicated in the raw data.
`Post#cooked`:
```
<p><a href=\"https://meta.discourse.org/some.png\" class=\"onebox\" target=\"_blank\" rel=\"nofollow noopener\">https://meta.discourse.org/some.png</a></p>
```
`PostSearchData#raw_data` Before:
```
This is a test topic 0 Uncategorized https://meta.discourse.org/some.png discourse org/some png https://meta.discourse.org/some.png discourse org/some png
```
`PostSearchData#raw_data` After:
```
This is a test topic 0 Uncategorized https://meta.discourse.org/some.png meta discourse org
```
2. Ligthbox being included in search pollutes the
`PostSearchData#raw_data` unncessarily.
From 28 March 2018 to 28 March 2019, searches for the term `image` on
`meta.discourse.org` had a click through rate of 2.1%. Non-lightboxed images are not included in indexing for search yet we were indexing content within a lightbox. Also, search for terms like `image` was affected we were using `Pasted image` as the filename for
uploads that were pasted.
`Post#cooked`
```
<p>Let me see how I can fix this image<br>\n<div class=\"lightbox-wrapper\"><a class=\"lightbox\" href=\"https://meta.discourse.org/some.png\" title=\"some.png\" rel=\"nofollow noopener\"><img src=\"https://meta.discourse.org/some.png\" width=\"275\" height=\"299\"><div class=\"meta\">\n<svg class=\"fa d-icon d-icon-far-image svg-icon\" aria-hidden=\"true\"><use xlink:href=\"#far-image\"></use></svg><span class=\"filename\">some.png</span><span class=\"informations\">1750×2000</span><svg class=\"fa d-icon d-icon-discourse-expand svg-icon\" aria-hidden=\"true\"><use xlink:href=\"#discourse-expand\"></use></svg>\n</div></a></div></p>
```
`PostSearchData#raw_data` Before:
```
This is a test topic 0 Uncategorized Let me see how I can fix this image some.png png https://meta.discourse.org/some.png discourse org/some png some.png png 1750×2000
```
`PostSearchData#raw_data` After:
```
This is a test topic 0 Uncategorized Let me see how I can fix this image
```
In terms of indexing performance, we now have to parse the given HTML
through nokogiri twice. However performance is not a huge worry here since a string length of 194170 takes only 30ms
to scrub plus the indexing takes place in a background job.
On smaller mobile devices, the height of the advanced search filters takes up
the whole real estate that it requires the user to scroll down
significantly in order to view the results.
Includes support for flags, reviewable users and queued posts, with REST API
backwards compatibility.
Co-Authored-By: romanrizzi <romanalejandro@gmail.com>
Co-Authored-By: jjaffeux <j.jaffeux@gmail.com>
Previously we relied on the provider name matching the name of the icon. Now icon names are explicitly set. Plugin providers which do not define an icon will get the default "sign-in-alt" icon
Improves usability of header search icon while user is already in full page search in mobile.
Currently, hitting search icon a second time empties input and does not scroll up to show search form.
This commit scrolls up to show form and sets focus on input.
- The test_email job is removed, because it was always being run synchronously (not in sidekiq)
- 34b29f62 added a bypass for critical emails, to match the spec. This removes the bypass, and removes the spec.
- This adapts the specs for 72ffabf6, so that they check for emails being sent
- This reimplements c2797921, allowing test emails to be sent even when emails are disabled
* Revert "FIX: discourse client should know about Logster (#7232)"
This reverts commit bfcbc4d2d6.
* FIX: discourse client should know about Logster (simpler approach)
- s3_force_path_style was added as a Minio specific url scheme but it has never been well supported in our code base.
- Our new migrate_to_s3 rake task does not work reliably with path style urls too
- Minio has also added support for virtual style requests i.e the same scheme as AWS S3/DO Spaces so we can rely on that instead of using path style requests.
- Add migration to drop s3_force_path_style from the site_settings table
* Check if user is a member of a group or if the group has members
this is used in the template to conditionally show the relevant markup
* Check if user is suspended or if they have a bio
this is used in the template to conditionally show the relevant markup
* Simplify group-card template
* Simplify user-card template
* Refactor user and group cards CSS
* Check if user is new or if user is staff
* Style fixes
- round avatar margin
- use a more standard margin for buttons
- adds lighter font color for new users
- makes some suspension text bold (used to be <b> tags in template)
- ensures images in group bio are responsive
* user card template fixes
- adds quotes to link href attributes
- wraps some strings in tags for more consistent styling
* group card fixes
- adds quotes to link href attributes
- fixes membership button login action
- wraps some strings with tags for consistent styling
* closure action fixes
* closure action fix
* uses core variables instead of new colors and removes unused styles
* Uses better property names
1. Remove `cooked` becaused only the blurb is used on the client side to
display search results.
2. Remove `ignored` because the result is not used in anyway when
searching for pots.
## Before
```
Acceptance: Composer::Image resizing buttons: 7985ms
Acceptance: Composer::Edit the first post: 3854ms
Acceptance: Composer::Composer with dirty edit can toggle to another edit: 3707ms
Acceptance: Composer::Composer can toggle between edit and reply: 3704ms
Acceptance: Composer::Tests the Composer controls: 3685ms
Acceptance: Composer::Composer draft with dirty reply can toggle to edit: 3670ms
Acceptance: Composer::Composer can toggle layouts (open, fullscreen and draft): 3278ms
Acceptance: Composer::Switching composer whisper state: 3266ms
Acceptance: Composer::Create a topic with server side errors: 3253ms
Acceptance: Composer::Composer with dirty reply can toggle to edit: 3189ms
Acceptance: Composer::Create a Topic: 3168ms
Acceptance: Composer::Create an enqueued Topic: 2767ms
Acceptance: Composer::Posting on a different topic: 2394ms
Acceptance: Composer::Composer can switch between edits: 2318ms
Acceptance: Composer::Create an enqueued Reply: 2317ms
Acceptance: Composer::Create a Reply: 2292ms
Acceptance: Composer::Checks for existing draft: 1696ms
Acceptance: Composer::Composer can toggle between reply and createTopic: 1415ms
Acceptance: Composer::Composer can toggle whispers: 1296ms
Acceptance: Composer::Loading draft also replaces the recipients: 594ms
```
## After
```
Acceptance: Composer::Composer can toggle layouts (open, fullscreen and draft): 2305ms
Acceptance: Composer::Switching composer whisper state: 2205ms
Acceptance: Composer::Composer draft with dirty reply can toggle to edit: 2185ms
Acceptance: Composer::Composer can toggle between edit and reply: 1719ms
Acceptance: Composer::Composer with dirty edit can toggle to another edit: 1682ms
Acceptance: Composer::Composer with dirty reply can toggle to edit: 1657ms
Acceptance: Composer::Composer can toggle between reply and createTopic: 1412ms
Acceptance: Composer::Posting on a different topic: 1341ms
Acceptance: Composer::Edit the first post: 1327ms
Acceptance: Composer::Create an enqueued Reply: 1306ms
Acceptance: Composer::Composer can toggle whispers: 1265ms
Acceptance: Composer::Composer can switch between edits: 1260ms
Acceptance: Composer::Create a Reply: 1259ms
Acceptance: Composer::Create a topic with server side errors: 1183ms
Acceptance: Composer::Checks for existing draft: 1172ms
Acceptance: Composer::Create a Topic: 1130ms
Acceptance: Composer::Image resizing buttons: 921ms
Acceptance: Composer::Create an enqueued Topic: 731ms
Acceptance: Composer::Tests the Composer controls: 654ms
Acceptance: Composer::Loading draft also replaces the recipients: 540ms
```
That is not a typo... image resizing button went from 8 seconds to 1 second
