Resolves https://codeberg.org/forgejo/forgejo/issues/10584
Codeberg mostly uses about 30 connections with rare peaks, so I think it's safe to assume that this value is reasonable even for large Forgejo deployments, and only needs to be adjusted for very busy instances. In practice, the open connections likely remain lower due to the connection lifetime and maximum idle connections (which could maybe also be tuned later). A value of 30 would technically allow 3 full-loaded Forgejo instances to operate on one PostgreSQL instance (30 * 3 = 90 < 97) without conflicting.
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10821
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Co-authored-by: Otto Richter <git@otto.splvs.net>
Co-committed-by: Otto Richter <git@otto.splvs.net>
* Add some test that only snapshot relevant content
* Allow adding marging around the element in case the environment is relevant (e.g. the location of an element relative to the parent, but excluding the environment of the parent)
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/9499
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Otto Richter <git@otto.splvs.net>
Co-committed-by: Otto Richter <git@otto.splvs.net>
In https://codeberg.org/forgejo/forgejo/pulls/7906#issuecomment-5511884, I noticed that the e2e tests were failing without obvious reasons. I was able to reproduce locally with Mobile Chrome, but the error doesn't make sense to me.
So I tried to downgrade playwright to the previous version, and it works fine. I actually suspect a bug in playwright, but I currently lack the capacity to reach out to upstream with a reproducer (I also found conversation with the playwright team a little difficult, unless you have absolutely convincing arguments that the flakiness you observe is really their fault, which is very hard to prove).
Tests pass with this version of playwright. In order to detect such cases earlier, I added a way to run all playwright tests (which I thought I had added with the changed files patch, but apparently forgot).
All tests are triggered by an explicit label (but only after firing a next event, because the testing workflows don't listen to label changes) and when the PR title contains "playwright", which should cover at least renovate dependency updates of playwright and the axe framework (both contain playwright).
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8245
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Reviewed-by: Beowulf <beowulf@beocode.eu>
Co-authored-by: Otto Richter <git@otto.splvs.net>
Co-committed-by: Otto Richter <git@otto.splvs.net>
- directly click email address
- show prohibited login state
## Testing
- visit admin panel
- visit a user
- click on "view details"
- check that email address is clickable and opens local email editor
- check that "suspended account" status is visible
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/7062
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Otto Richter <git@otto.splvs.net>
Co-committed-by: Otto Richter <git@otto.splvs.net>
Accessibility:
- improved semantic layout
- Fixes unlabelled input for custom pronouns. CC @hazy
- Adds labels to dropdowns.
- Shortens certain texts for less verbose screen reader outputs and
people with slow reading speed.
- Turned optional username rename helper text with low contrast into
"normal" help text.
UI/UX:
- Removes section about primary email which is no longer managed in the
profile section.
- Fixes section about primary email not displaying in user settings when notifications are
not available.
- Removes primary email display, because it is not actually a form
element here. (Alternatively, we could display it and link to the
account settings for managing the email)
Closes https://codeberg.org/forgejo/forgejo/issues/1120
- Adds labels to reaction and context menu.
- Fixes taborder in markdown combobox buttons. They are now only one
"tab" stop and can be navigated with arrow buttons and in the right
order (previously, it would skip the table button).
- Generates more verbose output for the reactio selectors to provide
content for users who cannot identify the meaning of these buttons
visually. Explicit aria-labels are now preferred over auto-generated
ones.
- screenshots and basic accessibility scan of collapsed and expanded sections
- the dropdowns do not pass the accessibility checks, but I haven't found an easy fix
- I manually confirmed the dropdown behaviour via orca and firefox, though
- separate template
- ensure correct labelling of elements
- drop additional required indicators for field that already have
browser semantics (the icon has colour contrast issues anyway),
especially as the first dropdown cannot be left empty
- create screenshots on tests that create a custom page
- skip screenshot on a page that uses the default page (it is added
automatically and already exists)
- replace a useless response wait and assert the correct page style
instead
Goals:
- speedup
- less flakiness
- best practices and more use
- documentation
config:
- sync ports in Makefile and playwright config
(otherwise, some tests fail locally because they assert the full URL including the (wrong) port)
- even more generous timeouts
- limit workers to one again (because I finally understand how
Playwright works)
- allow nested functions to group them together with the related test
all:
- deprecate waitForLoadState('networkidle')
- it is discouraged as per https://playwright.dev/docs/api/class-page#page-wait-for-load-state
- I could not find a usage that seems to require it actually (see
added documentation in README)
- adding an exception should be made explicitly
- it does not do what you might expect anyway in most cases
- only log in when necessary
webauthn:
- verify that login is possible after disabling key
- otherwise, the cleanup was not necessary after the previous refactor to create a fresh user each
issue-sidebar / WIP toggle:
- split into smaller chunks
- restore original state first
- add missed assertion to fix race condition (not waiting
before state was reached)
- explicitly toggle the state to detect mismatch earlier
issue-sidebar / labels:
- restore original state first
- better waiting for background request
- test label needs to be set and either present, not-needed or manual
- if manual test label is set, PR description needs to contain a heading
(defined by '#') starting with "Test" (e.g. "Test instructions",
"Testing" etc)
This change does not bring the desired benefit, because the button is
reinitialized from a Vue component once loaded, overriding this change
and causing a visual glitch upon page load (the button is moving its
position).
See the comment:
> this code must match the code in BranchTagSelector.vue
This button is also used in other places such as the release list where
the additional margin does not fit well. As such, this needs a new
solution.
This reverts commit 18cad9d342.
- detect changed files for the run
- let e2e files specify which related files they "watch"
- only run e2e tests based on pattern matching or when generic files
change
- fallback to full runs if env not specified
- supports glob patterns in testfiles
- only runs tests on changes
- always runs tests without specified patterns
tests(e2e): refactor global watch patterns
tests(e2e): add watch patterns to test files
ci: cache frontend build across jobs
ci: ensure caches are saved with zstd
work around https://github.com/actions/cache/issues/1169
ci: require unit tests for remote cacher
- prevents unnecessary runs in case the unit tests already fail
- starts the integration tests about 2 minutes earlier
- should give some overall speedup to the CI run, because the long integration tests are run and finish earlier, and the cacher tests should still usually finish in time
- does not save any computing resources, just provides quicker results when runners are not under high load
Change word order for issue comment actions
- An attempt to address https://codeberg.org/forgejo/forgejo/issues/2650
Org team permissions improvements
- consistency: added missing dot
- clarity: explain what external units mean
- use dedicated keys to explain the permissions.
- split in read/write permissions
- use explicit labels for accessibility
- ext_wiki.desc and ext_issues.desc are no longer in use.
- allow running with multiple workers (tested with up to four workers locally which
didn't show signs of flakiness)
- prevent race condition with webauthn tests (running them in parallel
on the same user could prevent another test from logging in)
- fix flakiness on CI action status (Chromium sometimes needs a long
time to fill the href field, firefox is always faster)
This reverts commit e8585eff5c.
- when the PR title has the maximum length, the WIP toggle switch does nothing
- work around this by slightly reducing the max input size (- 10 characters for eventually long prefixes)
- test WIP toggling edge case in playwright
fix(e2e): increase timeouts
A look at recent runs suggests they should be increased globally. The timeouts in the config file have no timeout by default.
- retrieved by the commit hash
- removes bindata tags from integration tests, because it does not seem
to be required
- due to the missing automatically generated data, the zstd tests fail
(they use repo data including node_modules (!) as input to the test,
there is no apparent reason for the size constants)
includes:
- easier repo declaration for playwright tests by @Gusted
- full backend build for pushing Git repos by @Gusted
- playwright testing (which fails with the current diff algorithm, but
passes with the new)
- disable eslint rule for conditional expect, because it defeats the
purpose (working around it would result in much more complex test code
in our cases)
- adds a header to indicate creating a new rule
- test that header is different between new and edit form
- consistently avoids colons in the form
- excludes some accessibility checks that require a global solution for
forms
UX/Translation changes:
- new teams: remove redundant tooltips that don't add meaningful information
- move general information to table fieldset
- new teams: rename "general" to "custom" access for clarity
- new teams: show labels beside options on mobile
Accessibility:
- semantic form elements allow easier navigation (fieldset, mostly)
- improve better labelling of new teams table
- fix accessibility scan issues
- TODO: the parts that "disable" form elements were not yet touched and
are not really accessible to screenreaders
Technical:
- replace two JavaScript solutions with one CSS standard
- implement a simpler grid (.simple-grid)
- simplify markup
- remove some webhook settings specific CSS
Testing:
- check more form content for accessibility issues
- but exclude tooltips from the scan :(
- reuse existing form tests from previous PR
Modifies forms:
- (new) org team
- (new) repo webhook
- (new) repo protected branch
The forms are not completely rewritten to semantic HTML yet. The focus
of this change was on standard elements, some custom solutions were left
untouched for now.
- swaps the order fo permission radio buttons as per https://codeberg.org/forgejo/forgejo/issues/4983
- uses fieldsets to group related inputs
- ensures consistent styling across forms
- can be improved later, e.g. using horizontal lines between sections
- fixes: previous font size of labels was smaller than the font size of the help text
- help text are now part of the label, clicking them now also activates the input
- drop unused CSS (no required checkboxes in grouped class remain)
- playwright testing:
- move login boilerplate to utils
- automated form accessibility checking
- allow defining the scope, because legacy parts of the forms are not yet accessible
- assert some CSS properties that should not be overriden
- the Makefile adjustment was necessary, because eslint scanned some internal files in the tests/e2e/reports directory
