* Add Stores and StoresContext and use it in MatrixChat and RoomView
Added a new kind of class:
- Add God object `Stores` which will hold refs to all known stores and the `MatrixClient`. This object is NOT a singleton.
- Add `StoresContext` to hold onto a ref of `Stores` for use inside components.
`StoresContext` is created via:
- Create `Stores` in `MatrixChat`, assigning the `MatrixClient` when we have one set. Currently sets the RVS to `RoomViewStore.instance`.
- Wrap `MatrixChat`s `render()` function in a `StoresContext.Provider` so it can be used anywhere.
`StoresContext` is currently only used in `RoomView` via the following changes:
- Remove the HOC, which redundantly set `mxClient` as a prop. We don't need this as `RoomView` was using the client from `this.context`.
- Change the type of context accepted from `MatrixClientContext` to `StoresContext`.
- Modify alllll the places where `this.context` is used to interact with the client and suffix `.client`.
- Modify places where we use `RoomViewStore.instance` and replace them with `this.context.roomViewStore`.
This makes `RoomView` use a non-global instance of RVS.
* Linting
* SDKContext and make client an optional constructor arg
* Move SDKContext to /src/contexts
* Inject all RVS deps
* Linting
* Remove reset calls; deep copy the INITIAL_STATE to avoid test pollution
* DI singletons used in RoomView; DI them in RoomView-test too
* Initial RoomViewStore.instance after all files are imported to avoid cyclical deps
* Lazily init stores to allow for circular dependencies
Rather than stores accepting a list of other stores in their constructors,
which doesn't work when A needs B and B needs A, make new-style stores simply
accept Stores. When a store needs another store, they access it via `Stores`
which then lazily constructs that store if it needs it. This breaks the
circular dependency at constructor time, without needing to introduce
wiring diagrams or any complex DI framework.
* Delete RoomViewStore.instance
Replaced with Stores.instance.roomViewStore
* Linting
* Move OverridableStores to test/TestStores
* Rejig how eager stores get made; don't automatically do it else tests break
* Linting
* Linting and review comments
* Fix new code to use Stores.instance
* s/Stores/SdkContextClass/g
* Update docs
* Remove unused imports
* Update src/stores/RoomViewStore.tsx
Co-authored-by: Michael Telatynski <7t3chguy@gmail.com>
* Remove empty c'tor to make sonar happy
Co-authored-by: Michael Telatynski <7t3chguy@gmail.com>
* refactor: convert RoomViewStore from flux Store to standard EventEmitter
Parts of a series of experimental changes to improve the design of stores.
* Use a gen5 store for RoomViewStore for now due to lock handling
* Revert "Use a gen5 store for RoomViewStore for now due to lock handling"
This reverts commit 1076af071d997d87b8ae0b0dcddfd1ae428665af.
* Add untilEmission and tweak untilDispatch; use it in RoomViewStore
* Add more RVS tests; remove custom room ID listener code and use EventEmitter
* Better comments
* Null guard `dis` as tests mock out `defaultDispatcher`
* Additional tests
* Add labs flag for sliding sync; add sliding_sync_proxy_url to config.json
* Disable the labs toggle if sliding_sync_proxy_url is not set
* Do validation checks on the sliding sync proxy URL before enabling it in Labs
* Enable sliding sync and add SlidingSyncManager
* Get room subscriptions working
* Hijack renderSublists in sliding sync mode
* Add support for sorting alphabetically/recency and room name filters
* Filter out tombstoned rooms; start adding show more logic
list ranges update but the UI doesn't
* update the UI when the list is updated
* bugfix: make sure the list sorts numerically
* Get invites transitioning correctly
* Force enable sliding sync and labs for now
* Linting
* Disable spotlight search
* Initial cypress plugins for Sliding Sync Proxy
* Use --rm when running Synapse in Docker for Cypress tests
* Update src/MatrixClientPeg.ts
Co-authored-by: Travis Ralston <travisr@matrix.org>
* Update src/components/views/rooms/RoomSublist.tsx
Co-authored-by: Travis Ralston <travisr@matrix.org>
* Update src/settings/controllers/SlidingSyncController.ts
Co-authored-by: Travis Ralston <travisr@matrix.org>
* Update src/components/views/rooms/RoomSublist.tsx
Co-authored-by: Travis Ralston <travisr@matrix.org>
* WIP add room searching to spotlight search
* Only read sliding sync results when there is a result, else use the local cache
* Use feature_sliding_sync not slidingSync
* Some review comments
* More review comments
* Use RoomViewStore to set room subscriptions
* Comment why any
* Update src/components/views/rooms/RoomSublist.tsx
Co-authored-by: Michael Telatynski <7t3chguy@gmail.com>
* Fix cypress docker abstraction
* Iterate sliding sync proxy support
* Stash mostly functional test
* Update sliding sync proxy image
* i18n
* Add support for spaces; use list ID -> index mappings
- Mappings are more reusable and easier to understand than
racing for index positions.
- Register for all spaces immediately on startup.
* When the active space is updated, update the list registration
* Set spaces filter in the correct place
* Skeleton placeholder whilst loading the space
* Filter out spaces from the room list
* Use the new txn_id promises
* Ensure we actually resolve list registrations
* Fixmatrix-org/sliding-sync#30: don't show tombstoned search results
* Remove unused imports
* Add SYNCV3_SECRET to proxy to ensure it starts up; correct aliases for SS test
* Add another basic sliding sync e2e test
* Unbreak netlify
* Add more logging for debugging duplicate rooms
* If sliding sync is enabled, always use the rooms result even if it's empty
* Drop-in copy of RoomListStore for sliding sync
* Remove conditionals from RoomListStore - we have SlidingRoomListStore now
* WIP SlidingRoomListStore
* Add most sliding sync logic to SlidingRoomListStore
Still lots of logic in RoomSublist. Broken things:
- Join count is wrong completely.
- No skeleton placeholder when switching spaces.
* Migrate joined count to SS RLS
* Reinstate the skeleton UI when the list is loading
* linting
* Add support for sticky rooms based on the currently active room
* Add a bunch of passing SS E2E tests; some WIP
* Unbreak build from git merge
* Suppress unread indicators in sliding sync mode
* Add regression test for https://github.com/matrix-org/sliding-sync/issues/28
* Add invite test flows; show the invite list
The refactor to SS RLS removed the invite list entirely.
* Remove show more click as it wasn't the bug
* Linting and i18n
* only enable SS by default on netlify
* Jest fixes; merge conflict fixes; remove debug logging; use right sort enum values
* Actually fix jest tests
* Add support for favourites and low priority
* Bump sliding sync version
* Update sliding sync labs to be user configurable
* delint
* To disable SS or change proxy URL the user has to log out
* Review comments
* Linting
* Apply suggestions from code review
Co-authored-by: Travis Ralston <travisr@matrix.org>
* Update src/stores/room-list/SlidingRoomListStore.ts
Co-authored-by: Travis Ralston <travisr@matrix.org>
* Review comments
* Add issue link for TODO markers
* Linting
* Apply suggestions from code review
Co-authored-by: Michael Telatynski <7t3chguy@gmail.com>
* More review comments
* More review comments
* stricter types
Co-authored-by: Michael Telatynski <7t3chguy@gmail.com>
Co-authored-by: Travis Ralston <travisr@matrix.org>
* Improve accessibility and testability of Tooltip
Adding a role to Tooltip was motivated by React Testing Library's
reliance on accessibility-related attributes to locate elements.
* Make the ReadyWatchingStore constructor safer
The ReadyWatchingStore constructor previously had a chance to
immediately call onReady, which was dangerous because it was potentially
calling the derived class's onReady at a point when the derived class
hadn't even finished construction yet. In normal usage, I guess this
never was a problem, but it was causing some of the tests I was writing
to crash. This is solved by separating out the onReady call into a start
method.
* Rename 1:1 call components to 'LegacyCall'
to reflect the fact that they're slated for removal, and to not clash
with the new Call code.
* Refactor VideoChannelStore into Call and CallStore
Call is an abstract class that currently only has a Jitsi
implementation, but this will make it easy to later add an Element Call
implementation.
* Remove WidgetReady, ClientReady, and ForceHangupCall hacks
These are no longer used by the new Jitsi call implementation, and can
be removed.
* yarn i18n
* Delete call map entries instead of inserting nulls
* Allow multiple active calls and consolidate call listeners
* Fix a race condition when creating a video room
* Un-hardcode the media device fallback labels
* Apply misc code review fixes
* yarn i18n
* Disconnect from calls more politely on logout
* Fix some strict mode errors
* Fix another updateRoom race condition
* Remove deprecated feature_communities_v2_prototypes
* Update _components
* i18n
* delint
* Cut out a bit more dead code
* Carve into legacy components
* Carve into mostly the room list code
* Carve into instances of "groupId"
* Carve out more of what comes up with "groups"
* Carve out some settings
* ignore related groups state
* Remove instances of spacesEnabled
* Fix some obvious issues
* Remove now-unused css
* Fix variable naming for legacy components
* Update i18n
* Misc cleanup from manual review
* Update snapshot for changed flag
* Appease linters
* rethemedex
* Remove now-unused AddressPickerDialog
* Make ConfirmUserActionDialog's member a required prop
* Remove useless override from RightPanelStore
* Remove extraneous CSS
* Update i18n
* Demo: "Communities are now Spaces" landing page
* Restore linkify for group IDs
* Demo: Dialog on click for communities->spaces notice
* i18n for demos
* i18n post-merge
* Update copy
* Appease the linter
* Post-merge cleanup
* Re-add spaces_learn_more_url to the new SdkConfig place
* Round 1 of post-merge fixes
* i18n
Co-authored-by: Michael Telatynski <7t3chguy@gmail.com>
Fixes https://github.com/vector-im/element-web/issues/14602
This has proven to be less useful month over month, so it's time for it to go. The initial goal for keeping it was to track stability in the ~6 months post-release, which means it should have left around the start of 2021. At the time, Spaces was on the horizon with room list corruption issues though, so it was put on hold. These days, it's rock solid and doesn't appear to be causing issues. As such, the logging can leave in favour of more specific/targeted logging when we need it.
This might even speed things up as it's not hitting the SettingsStore every 0.2 seconds.
Fixes https://github.com/vector-im/element-web/issues/16799
This change replaces the "relative priority" system for filters with a kind model. The kind is used to differentiate and optimize when/where a filter condition is applied, resulting in a more stable ordering of the room list. The included documentation describes what this means in detail.
This also introduces a way to inhibit updates being emitted from the Algorithm class given what we're doing to the poor thing will cause it to do a bunch of recalculation. Inhibiting the update and implicitly applying it (as part of our updateFn.mark()/trigger steps) results in much better performance.
This has been tested on my own account with both communities and spaces of varying complexity: it feels faster, though the measurements appear to be within an error tolerance of each other (read: there's no performance impact of this).
Fixes https://github.com/vector-im/element-web/issues/15559
This isn't exactly perfect as an implementation: if the user refreshes immediately after forgetting then there is a good chance the room re-appears because of the sync accumulator. At the very least this change makes it so in *most* cases the room goes away, which is probably good enough until https://github.com/vector-im/element-web/issues/14038 can be implemented properly.
Does a thirdparty protocol lookup to the homeserver to get the
corresponding native/virtual user for a matrix ID. Stores the
mappings in room account data. Involves some slightly nasty workarounds
for that fact that room account data has no local echo.
