The script was relying on the external git-flow tool and on a local
`git config gitflow.*` initialisation, which made its behaviour depend on
the machine setup and hid what was actually done to the branches.
The release branch is now created with `git checkout -b`, and the release
is finished with explicit merge, tag, back-merge and branch deletion
commands. The merge and tag messages are provided explicitly so that the
generated history stays identical to the previous releases. Each step now
aborts the script with a clear message when it fails, instead of letting
the script continue.
Note that tools/release/release.sh (V1) still uses git flow.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The PR metadata checks move to a new PR Checks workflow, and the lint,
ktlint and detekt reports are now surfaced as inline annotations from the
quality workflow. This removes the danger-js dependency, its ad-hoc npm
installs and the need for the ElementBot token, so the checks also run on
pull requests from forks.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`git flow release finish` opened the editor three times, and the content was
never edited: twice for the merge commits, where the message generated by git is
always kept, and once for the message of the annotated tag, where a placeholder
was typed by hand. The last tags carry the message "tag" for that reason.
GIT_MERGE_AUTOEDIT=no keeps the merge messages git already generates, and -m
provides the tag message. git flow appends the tag name to that message, so the
tag of the next release will read "Release v26.09.0".
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
releaseV2.sh is a copy of release.sh in which the GitHub CLI is mandatory and
does what was previously done by hand: it finds and waits for the release.yml
run, downloads and extracts its artifacts, creates the pre-release with its 6
assets and the generated release notes, and prepends those notes to CHANGES.md.
It also compares the version code and version name of the signed artifacts to
the expected values instead of asking the user to check them, and no longer
needs ELEMENT_GITHUB_TOKEN, since the CLI uses its own credentials. release.sh
is left untouched, so that the next release can be done with releaseV2.sh while
keeping a way back if something goes wrong. The installation documentation is
also fixed, as it referred to an asset name that does not exist anymore.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The gallery gave no indication of how old the screenshots on the page were,
which is one of the remaining items on the GitHub Pages improvement list.
The generator now writes the time it ran alongside the screenshot data, and the
page renders it in the footer next to the existing links.
The 174 commits develop gained since this branch was cut touch preview data
in three ways, and each is resolved so the flipped konsist rule still holds
across the merged tree.
Two screens the branch had renamed were deleted upstream by "Remove the
account provider search screen and homeserver resolver": the
changeaccountprovider and searchaccountprovider packages. Their deletion
wins, so ChangeAccountProviderStatePreviewParam, SearchAccountProviderStatePreviewParam
and the two views that used them go with them rather than coming back
under the new name.
Two preview classes develop added still carried the old suffix and are
renamed here, along with the `@PreviewParameter` reference in each view:
GlobalSearchStateProvider -> GlobalSearchStatePreviewParam and
SendFailureDialogStateProvider -> SendFailureDialogStatePreviewParam.
Everything else merged textually. No production class extending
PreviewParameterProvider is left with a `Provider` name, and no reference
to any of the 194 renamed names survives outside the CHANGES.md entry that
records a past pull request title.
The `Provider` suffix reads as a dependency provider rather than as preview
data, and it collides with the real providers in the codebase — `StringProvider`,
`PickerProvider`, the login feature's `AccountProvider` — so the name says
nothing about which of the two a class is. `PreviewParam` says it.
195 classes are renamed, along with the 185 files named after them, the konsist
rule that enforces the convention, the presentation-class file templates and the
onboarding documentation.
This cannot be split. `Konsist.scopeFromProduction()` scans the whole project, so
flipping the assertion before every class has moved leaves the rule red, and a
rule that accepted either suffix during a transition would be exactly the kind of
compatibility layer this convention change is meant to remove.
Two judgement calls in the rule itself. The `name.endsWith("PreviewProvider")`
exclusion is dropped: its only job was to reject a redundant `Preview` before
`Provider`, and the new suffix already contains it; no class currently ends with
`PreviewProvider`, so removing it changes nothing today. And
`EditableAvatarViewUriProvider` is renamed rather than deleted from the exclusion
list even though no such class exists in this repository, since it presumably
comes from the modules overlaid for the enterprise build.
Nothing keyed off the old suffix: no snapshot filename contains a provider class
name, so no Paparazzi image changes, and the Kover variants match on `*Presenter`,
`*State` and `*ViewKt`, so coverage bounds are unaffected.
fix-pg-map-id.py rewrites the pg_map_id of a dex file, but the dex
format changed and its regex no longer matches anything, so the step is
a no-op and the F-Droid build is reproducible without it. inplace-fix.py
only existed to drive it; none of its other commands are used here.
The removed call also re-ran zipalign -P 16 on each APK. AGP aligns the
release APKs at packaging time and apksigner is already invoked with
--alignment-preserved, so the signed output stays 16 KB aligned.
Generating a feature with the Generate Module from Template plugin produced a
module that could not build: features/<name>/api/build.gradle.kts was written
from the impl template, giving the api module a dependency on itself and an
impl namespace.
FeatureModule.json declared BUILD_GRADLE_API and BUILD_GRADLE_IMPL, both
defaulting to build.gradle.kts, and used them as fileTemplates keys. The plugin
resolves keys to filenames before matching, so both collapsed to one key and
the second overwrote the first. The same matching code falls back to a suffix
match on the path, so DefaultXEntryPoint also matched the key XEntryPoint and
was generated from the api template.
Key both by directory, which the plugin supports for exactly this case, and
drop the two now-unused placeholders. The file templates themselves were
correct and are unchanged. The onboarding doc gains a note that an already
imported template must be removed before re-importing, since the plugin keeps
its own copy.
* Remove no longer temporary string
These are now in localazy and the sync strings CI flow was failing because we have 2 conflicting values in `temporary.xml` and `localazy.xml`
* Use the right placeholder value in `screen_local_network_opt_in_subtitle`
* Localazy: improve search, parallelize checks
The search for XML files has been narrowed down and it's several times faster now, and processing the files is now done in parallel
* Improve script to also find and delete previous files
Also add the option of having a `-v|--verbose` flag to print which files we're modifying
* Use `LocalBuildMeta` to get the app name inside `LocalNetworkPermissionDialogView`
* Allow finding all the forbidden terms in the strings and translations and then return an error exit code instead of a success one
---------
Co-authored-by: ElementBot <android@element.io>
* Simplify how event content with captions are displayed
The caption component will now be added in `TimelineItemEventContentView` instead of separately in each `TimelineItem__View` component
* Make sure `CoilMediaFetcher` methods to fetch media throw errors
We need to catch them later to check if the content scanner marked some thumbnails or medias as invalid/unsafe in `AsyncImage` instances
* Add a `ContentScannerService` that allows us to manually scan a list of medias in an event and then update and store the results
The results take the form of `ContentValidationValue`, stored in a `ContentValidationState` instance tied to the event
* Create `EventContentValidationCache` to keep the `ContentValidationStates` for events in a room
* Expose `EventContentValidationCache` to Compose using `LocalEventContentValidationState`
Also add a couple of helper functions to avoid having to manually remember the states
* Use the new components to display invalid content in the timeline items
* Add `blurhash` to most media items so we can use it while validating the thumbnails/medias
* Fix custom layout issues with screenshot and UI tests
* Also check the validation state for replied-to events in the composer
* Fix and add tests
* Handle unrecoverable content validation errors too: display a 'not found' UI
* Improve the performance of `EqualWidthColumn` for the single item case
* Update screenshots
---------
Co-authored-by: ElementBot <android@element.io>
* Collapse runs of deleted messages in the timeline (#7111)
* Collapse runs of deleted messages in the timeline
Collapse a run of three or more consecutive deleted messages into a single
expandable group, reusing the existing state-change grouping (the same way
element-web does it) instead of showing one "Message removed" placeholder per
deletion. The header shows the count only ("N deleted messages"); the Rust SDK
does not expose who performed the redaction, so we don't attribute it.
On by default. Runs shorter than three are left as individual placeholders, and
day dividers and read receipts are preserved (a run is broken by any
non-redacted item, so a day is never emptied).
Run the collapse as the final step of TimelineItemGrouper.group() instead of
a separate pass in the presenter, as suggested in review. Kept as its own
finalization step rather than via canBeGrouped() so the group stays all
redacted and the count-only header is preserved.
The group id is resolved through the grouper's existing id registry, like the
other groups, so it stays stable however the run grows and the user's
expand/collapse state is kept across timeline updates.
* Replace temporary strings with final ones
* Update screenshots
---------
Co-authored-by: Hi Dude! <klarkmonty@tuta.io>
Co-authored-by: manfrommedan <manfrommedan@users.noreply.github.com>
Co-authored-by: ElementBot <android@element.io>
Benoit Marty
Benoit Marty
ganfra
Jorge Martin Espinosa
jmartinesp
hayaksi1
Benoit Marty
Richard van der Hoff
ganfra