* WIP doodles on MembershipManager test cases
* .
* initial membership manager test setup.
* Updates from discussion
* revert renaming comments
* remove unused import
* fix leave delayed event resend test.
It was missing a flush.
* comment out and remove unused variables
* es lint
* use jsdom instead of node test environment
* remove unused variables
* remove unused export
* temp
* review
* fixup tests
* more review
* remove wait for expect dependency
* temp
* fix wrong mocked meberhsip template
* rename MembershipManager -> LegacyMembershipManager
And remove the IMembershipManager from it
* Add new memberhsip manager
* fix tests to be compatible with old and new membership manager
* Comment cleanup
* Allow join to throw
- Add tests for throwing cases
- Fixs based on tests
* introduce membershipExpiryTimeoutSlack
* more detailed comments and cleanup
* warn if slack is misconfigured and use default values instead
* fix action resets.
* flatten MembershipManager.spec.ts
* rename testEnvironment to memberManagerTestEnvironment
* allow configuring Legacy manager in the matrixRTC session
* deprecate LegacyMembershipManager
* remove usage of waitForExpect
* flatten tests and add comments
* clean up leave logic branch
* add more leave test cases
* use defer
* review ("Some minor tidying things for now.")
* add onError for join method and cleanup
* use pop instead of filter
* fixes
* simplify error handling and MembershipAction
Only use one membership action enum
* Add diagram
* fix new error api in rtc session
* fix up retry counter
* fix lints
* make unrecoverable errors more explicit
* fix tests
* Allow multiple retries on the rtc state event http requests.
* use then catch for startup
* no try catch 1
* update expire headroom logic
transition from try catch to .then .catch
* replace flushPromise with advanceTimersByTimeAsync
* fix leaving special cases
* more unrecoverable errors special cases
* move to MatrixRTCSessionManager logger
* add state reset and add another unhandleable error
The error occurs if we want to cancel the delayed event we still have an id for but get a non expected error.
* missed review fixes
* remove @jest/environment dependency
* Cleanup awaits and Make mock types more correct.
Make every mock return a Promise if the real implementation does return a pormise.
* remove flush promise dependency
* fix not recreating default state on reset
This broke all tests since we only created the state once and than passed by ref
* Use per action rate limit and retry counter
There can be multiple retries at once so we need to store counters per action
e.g. the send update membership and the restart delayed could be rate limited at the same time.
* add linting to matrixrtc tests
* Add fix async lints and use matrix rtc logger for test environment.
* prettier
* review step 1
* change to MatrixRTCSession logger
* review step 2
* make LoopHandler Private
* update config to use NewManager wording
* emit error on rtc session if the membership manager encounters one
* network error and throw refactor
* make accessing the full room deprecated
* remove deprecated usage of full room
* Clean up the deprecation
* add network error handler and cleanup
* better logging, another test, make maximumNetworkErrorRetryCount configurable
* more logging & refactor leave promise
* add ConnectionError as possible retry cause
* Make it work in embedded mode with a server that does not support delayed events
* review iteration 1
* review iteration 2
* first step in improving widget error handling
* make the embedded client throw ConnectionErrors where desired.
* fix tests
* delayed event sending widget mode stop gap fix.
* improve comment
* fix unrecoverable error joinState (and add JoinStateChanged) emission.
* check that we do not add multipe sendFirstDelayed Events
* also check insertions queue
* always log "Missing own membership: force re-join"
* Do not update the membership if we are in any (a later) state of sending our own state.
The scheduled states MembershipActionType.SendFirstDelayedEvent and MembershipActionType.SendJoinEvent both imply that we are already trying to send our own membership state event.
* make leave reset actually stop the manager.
The reset case was not covered properly. There are cases where it is not allowed to add additional events after a reset and cases where we want to add more events after the reset. We need to allow this as a reset property.
* fix tests (and implementation)
* Allow MembershipManger to be set at runtime via JoinConfig.membershipManagerFactory
* Map actions into status as a sanity check
* Log status change after applying actions
* Add todo
* Cleanup
* Log transition from earlier status
* remove redundant status implementation
also add TODO comment to not forget about this.
* More cleanup
* Consider insertions in status()
* Log duration for emitting MatrixRTCSessionEvent.MembershipsChanged
* add another valid condition for connected
* some TODO cleanup
* review add warning when using addAction while the scheduler is not running.
* es lint
* refactor to return based handler approach (remove insertions array)
* refactor: Move action scheduler
* refactor: move different handler cases into separate functions
* linter
* review: delayed events endpoint error
* review
* Suggestions from pair review
* resetState is actually only used internally
* Revert "resetState is actually only used internally"
This reverts commit 6af4730919ec07ce9aaad8de35c27ac6b98a3019.
* refactor: running is part of the scheduler (not state)
* refactor: move everything state related from schduler to manager.
* review
* Update src/matrixrtc/NewMembershipManager.ts
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* review
* public -> private + missed review fiexes (comment typos)
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@matrix.org>
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* WIP doodles on MembershipManager test cases
* .
* initial membership manager test setup.
* Updates from discussion
* revert renaming comments
* remove unused import
* fix leave delayed event resend test.
It was missing a flush.
* comment out and remove unused variables
* es lint
* use jsdom instead of node test environment
* remove unused variables
* remove unused export
* temp
* review
* fixup tests
* more review
* remove wait for expect dependency
* flatten tests and add comments
* add more leave test cases
* use defer
* remove @jest/environment dependency
* Cleanup awaits and Make mock types more correct.
Make every mock return a Promise if the real implementation does return a pormise.
* remove flush promise dependency
* add linting to matrixrtc tests
* Add fix async lints and use matrix rtc logger for test environment.
* prettier
* change to MatrixRTCSession logger
* make accessing the full room deprecated
* remove deprecated usage of full room
* Clean up the deprecation
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@matrix.org>
* update join and leave internal api.
* rename onMembershipUpdate and triggerCallMembershipEventUpdate to onMembershipsUpdate
This makes it more clear that we do not talk about our own membership but all memberships in the session
* cleanup MembershipManager
- add comments and interface how to test this class.
- sort methods by public/private
- make triggerCallMembershipEventUpdate private
* docstrings for getFocusInUse and getActiveFocus
* simplify tests and make them only use MembershipManagerInterface methods.
This allows to exchange the membershipManager with a different implementation.
* convert interface to abstract class.
* review (implement interface, make interface internal, dont change public api.)
* Make the interface an actual interface.
The actual constructor of the class now contains the `Pick` to define what it needs from the client.
* move update condition into MembershipManager
* renaming public api
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* split joinConfig
- myMembership related properties get moved into its own interface
* Add MyMembershipManager
* Remove methods and functions that are from MatrixRTCSession (they now live in MyMembershipManager)
* Refactor MatrixRTCSession to use myMembershipManager
* fix tests
* review
* get rid of more memberhsip manager usage in tests
* review - fix tests using private membershipManager props
* fix circular import
* remove all legacy call related code and adjust tests.
We actually had a bit of tests just for legacy and not for session events. All those tests got ported over so we do not remove any tests.
* dont adjust tests but remove legacy tests
* Remove deprecated CallMembership.getLocalExpiry()
* Remove references to legacy in test case names
* Clean up SessionMembershipData tsdoc
* Remove CallMembership.expires
* Use correct expire duration.
* make expiration methods not return optional values and update docstring
* add docs to `SessionMembershipData`
* Add new tests for session type member events that before only existed for legacy member events.
This reverts commit 795a3cffb61d672941c49e8139eb1d7b15c87d73.
* remove code we do not need yet.
* Cleanup
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@matrix.org>
* remove all legacy call related code and adjust tests.
We actually had a bit of tests just for legacy and not for session events. All those tests got ported over so we do not remove any tests.
* dont adjust tests but remove legacy tests
* Remove deprecated CallMembership.getLocalExpiry()
* Remove references to legacy in test case names
* Clean up SessionMembershipData tsdoc
* Remove CallMembership.expires
* Use correct expire duration.
* make expiration methods not return optional values and update docstring
* add docs to `SessionMembershipData`
* Use `MSC4143` (instaed of `non-legacy`) wording in comment
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* Incorporate feedback from review
* Fix test name
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@matrix.org>
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* Set retry counts of event updating to 1000 (from 1)
With it being set to one the following issue could occur:
```
// If sending state cancels your own delayed state, prepare another delayed state
// TODO: Remove this once MSC4140 is stable & doesn't cancel own delayed state
if (this.disconnectDelayId !== undefined) {
try {
const knownDisconnectDelayId = this.disconnectDelayId;
await resendIfRateLimited(
() =>
this.client._unstable_updateDelayedEvent(
knownDisconnectDelayId,
UpdateDelayedEventAction.Restart,
),
1000,
);
} catch (e) {
logger.warn("Failed to update delayed disconnection event, prepare it again:", e);
this.disconnectDelayId = undefined;
await prepareDelayedDisconnection();
}
}
```
This code looks like the `catch(e)` could never be triggered with 429 (rate limit) because they would be caught by `await resendIfRateLimited`. EXCEPT that this is only happening once: `resendIfRateLimited<T>(func: () => Promise<T>, numRetriesAllowed: number = 1)`. So as soon as the server sends two rate limits in a row we get the following:
- we get into the `catch(e)` because of the rate limit
- we forget about `this.disconnectDelayId = undefined`
- we start a new delayed event `await prepareDelayedDisconnection();`
- we do not anymore update the old delayed event which is still running!
- the running delay event will make us disconnect from the call (call member becomes `{}`)
- we get into our outher error catching mechanism that resends the new state event
- this cancels the newly created delay leave event (`await prepareDelayedDisconnection();`)
- and create another delay leave event.
- but if we are still reate limited (chances are really high due to the reconnect), this loop will REPEAT
* also check for M_NOT_FOUND
* Leave retry at current level
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@matrix.org>
* Handle M_MAX_DELAY_EXCEEDED errors
Use a lower delay time if the server rejects a delay as too long.
* Add test
* Lint test
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Robin <robin@robin.town>
* Test computed expiry timeout value
---------
Co-authored-by: Robin <robin@robin.town>
* When state says you've left ongoing call, rejoin
When receiving a state change that says you are no longer a member of a
RTC session that you are actually still participating in, send another
state event to put yourself back in the session membership.
This can happen when an administrator overwrites your call membership
event (which is allowed even with MSC3757's restrictions on state), or
if your delayed disconnection event (via MSC4140) timed out before your
client could send a heartbeat to delay it further.
* Don't emit state changed on join recovery
* MatrixRTCSession: handle rate limit errors
* Lint
* Handle ratelimiting for non-legacy state setting
Each request must be retried, as the non-legacy flow involves a sequence
of requests that must resolve in order.
* Fix broken test
* Check for MSC3757 instead of the unmerged MSC3779
* Move helper out of beforeEach
* Test ratelimit errors
* Include HTTP response headers in MatrixError
* Lint
* Support MSC4041 / Retry-After header
* Fix tests
* Remove redundant MatrixError parameter properties
They are inherited from HTTPError, so there is no need to mark them as
parameter properties.
* Comment that retry_after_ms is deprecated
* Properly handle colons in XHR header values
Also remove the negation in the if-condition for better readability
* Improve Retry-After parsing and docstring
* Revert ternary operator to if statements
for readability
* Reuse resolved Headers for Content-Type parsing
* Treat empty Content-Type differently from null
* Add MatrixError#isRateLimitError
This is separate from MatrixError#getRetryAfterMs because it's possible
for a rate-limit error to have no Retry-After time, and having separate
methods to check each makes that more clear.
* Ignore HTTP status code when getting Retry-After
because status codes other than 429 may have Retry-After
* Catch Retry-After parsing errors
* Add test coverage for HTTP error headers
* Update license years
* Move safe Retry-After lookup to global function
so it can more conveniently check if an error is a MatrixError
* Lint
* Inline Retry-After header value parsing
as it is only used in one place and doesn't need to be exported
* Update docstrings
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* Use bare catch
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* Give HTTPError methods for rate-limit checks
and make MatrixError inherit them
* Cover undefined errcode in rate-limit check
* Update safeGetRetryAfterMs docstring
Be explicit that errors that don't look like rate-limiting errors will
not pull a retry delay value from the error.
* Use rate-limit helper functions in more places
* Group the header tests
---------
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* Do not rotate MatrixRTC media encryption key when a new member joins a call
This change reverts https://github.com/matrix-org/matrix-js-sdk/pull/4422.
Instead, the rotation when a new member joins will be reintroduced as part of supporting to-device based MatrixRTC encryption key distribution.
* Improve function name
* Refactor/simplify Promises in MatrixRTCSession
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* Fix+document+test leaveRoomSession's return value
* Throw instead of using expect in teardown
because lint rules forbid using expect outside of test functions
---------
Co-authored-by: Hugh Nimmo-Smith <hughns@users.noreply.github.com>
* Prepare delayed call leave events more reliably
- Try sending call join after preparing delayed leave
- On leave, send delayed leave instead of a new event
* Don't rely on errcodes for retry logic
because they are unavailable in widget mode
* Make arrow method readonly
SonarCloud rule typescript:S2933
* Test coverage for restarting delayed call leave
* Remove unneeded unstable_features mock
It's unneeded because all affected methods are mocked
* Don't share full key history for RTC per-participant encryption
Also record stats for how many keys have been sent/received and age of those received
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Robin <robin@robin.town>
* Add comment about why we track total age of events
---------
Co-authored-by: Robin <robin@robin.town>
* Handle MatrixRTC encryption keys arriving out of order
* Apply suggestions from code review
Co-authored-by: Andrew Ferrazzutti <andrewf@element.io>
* Suggestion from code review
---------
Co-authored-by: Andrew Ferrazzutti <andrewf@element.io>
* Resend MatrixRTC encryption keys if a membership has changed
* JSDoc
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Andrew Ferrazzutti <andrewf@element.io>
* Add note about using Set. symmetricDifference() when available
* Always store latest fingerprints
Should reduce unnecessary retransmits
* Refactor
---------
Co-authored-by: Andrew Ferrazzutti <andrewf@element.io>
Akin to how legacy call events should be sent in rooms where there is
any ongoing legacy call, send non-legacy events in rooms where there are
only non-legacy calls; else fall back to the config preference.
and use them for more reliable MatrixRTC session membership events.
Also implement "parent" delayed events, which were in a previous version
of the MSC and may be reintroduced or be part of a new MSC later.
NOTE: Still missing is support for sending encrypted delayed events.
* Be specific about what is considered a MSC4143 call member event.
* review
* check for empty event first
* Optimize for new session type events
If its a session type event we do not want to run two "key in" checks. We expect legacy events to be the less comment type going forward.
* awkward but necessary key count optimization
* Prefix the user+device state key if needed
No need to prefix it for rooms that use MSC3779.
Otherwise, prefix it to bypass the auth rule for state events with keys
starting with @.
* Use RegExp.exec() method instead
Sonar typescript:S6594
* Split nested ternary operator into method
Sonar typescript:S3358
* Add test coverage
* Use legacy call membership if anyone else is
* Convert nullish to boolean
* Update tests
* Lint
* Use computed decision to use legacy events or not
* Check if discovered legacy sessions are ongoing
* Lint
* Lint again
* Increase test coverage
* Refactor to preferred and active foci.
Signed-off-by: Timo K <toger5@hotmail.de>
* make the sdk compatible with MSC4143 but still be backwards compatible
* comment fixes
* also fallback to legacy if the current member event is legacy
* use XOR types
* use EitherAnd
* make livekit Foucs types simpler
* review
* fix tests
* test work
* more review + more tests
* remove unnecassary await that is in conflict with the comment
* make joinRoomSession sync
* Update src/matrixrtc/MatrixRTCSession.ts
Co-authored-by: Andrew Ferrazzutti <af_0_af@hotmail.com>
* review
* fix
* test
* review
* review
* comment clarification
* typo
---------
Signed-off-by: Timo K <toger5@hotmail.de>
Co-authored-by: Andrew Ferrazzutti <af_0_af@hotmail.com>
* Introduce Membership TS type
* Adapt the Membership TS type to be an enum
* Add docstrings for KnownMembership and Membership
* Move Membership types into a separate file, exported from types.ts
---------
Co-authored-by: Stanislav Demydiuk <s.demydiuk@gmail.com>
* ignore memberships of users that are not in the call
Signed-off-by: Timo K <toger5@hotmail.de>
* recompute memberships on room member change.
Signed-off-by: Timo K <toger5@hotmail.de>
* fix Tests and add test for left member
Signed-off-by: Timo K <toger5@hotmail.de>
* fix event type
Signed-off-by: Timo K <toger5@hotmail.de>
* fix import desaster
Signed-off-by: Timo K <toger5@hotmail.de>
* fix mocks
Signed-off-by: Timo K <toger5@hotmail.de>
---------
Signed-off-by: Timo K <toger5@hotmail.de>
* add expire_ts compatibility to matrixRTC
Signed-off-by: Timo K <toger5@hotmail.de>
* add expire_ts
Signed-off-by: Timo K <toger5@hotmail.de>
* rename expire_ts -> expires_ts
Signed-off-by: Timo K <toger5@hotmail.de>
* allow events without `expires`
Signed-off-by: Timo K <toger5@hotmail.de>
* fix test for expires_ts
Signed-off-by: Timo K <toger5@hotmail.de>
* comment clarification
Signed-off-by: Timo K <toger5@hotmail.de>
* add comment where one needs to use the origin_server_ts
Signed-off-by: Timo K <toger5@hotmail.de>
* add additional expires_ts tests
Signed-off-by: Timo K <toger5@hotmail.de>
* fix fake timer
Signed-off-by: Timo K <toger5@hotmail.de>
* change priority order to favor expires
Signed-off-by: Timo K <toger5@hotmail.de>
---------
Signed-off-by: Timo K <toger5@hotmail.de>
* Add CallNotifyEvent to support matrix rtc ringing
Signed-off-by: Timo K <toger5@hotmail.de>
* test SessionId
Signed-off-by: Timo K <toger5@hotmail.de>
* docs + sessionId->callId
Signed-off-by: Timo K <toger5@hotmail.de>
---------
Signed-off-by: Timo K <toger5@hotmail.de>
* WIP refactor for removing m.call events
* Always remember rtcsessions since we need to only have one instance
* Fix tests
* Fix import loop
* Fix more cyclic imports & tests
* Test session joining
* Attempt to make tests happy
* Always leave calls in the tests to clean up
* comment + desperate attempt to work out what's failing
* More test debugging
* Okay, so these ones are fine?
* Stop more timers and hopefully have happy tests
* Test no rejoin
* Test malformed m.call.member events
* Test event emitting
and also move some code to a more sensible place in the file
* Test getActiveFoci()
* Test event emitting (and also fix it)
* Test membership updating & pruning on join
* Test getOldestMembership()
* Test member event renewal
* Don't start the rtc manager until the client has synced
Then we can initialise from the state once it's completed.
* Fix type
* Remove listeners added in constructor
* Stop the client here too
* Stop the client here also also
* ARGH. Disable tests to work out which one is causing the exception
* Disable everything
* Re-jig to avoid setting listeners in the constructor
and re-enable tests
* No need to rename this anymore
* argh, remove the right listener
* Is it this test???
* Re-enable some tests
* Try mocking getRooms to return something valid
* Re-enable other tests
* Give up trying to get the tests to work sensibly and deal with getRooms() returning nothing
* Oops, don't enable the ones that were skipped before
* One more try at the sensible way
* Didn't work, go back to the hack way.
* Log when we manage to send the member event update
* Support `getOpenIdToken()` in embedded mode (#3676)
* Call `sendContentLoaded()` (#3677)
* Start MatrixRTC in embedded mode (#3679)
* Reschedule the membership event check
* Bump widget api version
* Add mock for sendContentLoaded()
* Embeded mode pre-requisites
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Embeded mode E2EE
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Encryption condition
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Revert "Embeded mode pre-requisites"
This reverts commit 8cd73702052609c995ad754e31f85d0da0be4aa9.
* Get back event type
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
fds
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Change embedded E2EE implementation
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* More log detail
* Fix tests
and also better assert because the tests were passing undefined which
was considered fine because we were only checking for null.
* Simplify updateCallMembershipEvent a bit
* Split up updateCallMembershipEvent some more
* Use `crypto.getRandomValues()`
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Rename to `membershipToUserAndDeviceId()`
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Better error
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Add log line
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Add comment
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Send call ID in enc events
(also a small refactor)
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Revert making `joinRoomSession()` async
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Make `client` `private` again
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Just use `toString()`
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix `callId` check
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix map
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix map compare
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix emitting
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Explicit logging
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Refactor
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Make `updateEncryptionKeyEvent()` public
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Only update keys based on others
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix call order
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Improve logging
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Avoid races
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Revert "Avoid races"
This reverts commit f65ed72d6e.
* Add try-catch
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Make `updateEncryptionKeyEvent()` private
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Handle indices and throttling
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix merge mistakes
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Mort post-merge fixes
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Split out key generation from key sending
And send all keys in a key event (changes the format of the key event)
rather than just the one we just generated.
* Remember and clear the timeout for the send key event
So we don't schedule more key updates if one is already pending.
Also don't update the last sent time when we didn't actually send the
keys.
* Make key event resends more robust
* Attempt to make tests pass
* crypto wasn't defined at all
* Hopefully get interface right
* Fix key format on the wire to base64
* Add comment
* More standard method order
* Rename encryptMedia
The js-sdk doesn't do media and therefore doesn't do media encryption
* Stop logging encryption keys now
* Use regular base64
It's not going in a URL, so no need
* Re-add base64url
randomstring was using it. Also give it a test.
* Add tests for randomstring
* Switch between either browser or node crypto
Let's see if this will work...
* Obviously crypto has already solved this
* Some tests for MatrixRTCSession key stuff
* Test keys object contents
* Change keys event format
To move away from m. keys
* Test key event retries
* Test onCallEncryption
* Test event sending & spam prevention
* Test event cancelation
* Test onCallEncryption called
* Better before/after member comparison
Only trigger for when members actually join, and just generally
make it a bit more understandable.
* Rotate per-participant keys when a member leaves
With a delay borth before making a new key, to try to batch up multiple
people leaving into a single key change, and a delay before actually
using the new key to allow time for it to arrive.
This increasingly feels like storing our own sender key in the same set
is suboptimal because we're starting to have to treat it more & more
specially.
* Some errors didn't have data
* Fix binary key comparison
& add log line
* Fix compare function with undefined values
* Test key rotation
* Test caught a merge bug!
* The missing word was, 'delay'
* More input validation
---------
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
Co-authored-by: Šimon Brandner <simon.bra.ag@gmail.com>
* WIP refactor for removing m.call events
* Always remember rtcsessions since we need to only have one instance
* Fix tests
* Fix import loop
* Fix more cyclic imports & tests
* Test session joining
* Attempt to make tests happy
* Always leave calls in the tests to clean up
* comment + desperate attempt to work out what's failing
* More test debugging
* Okay, so these ones are fine?
* Stop more timers and hopefully have happy tests
* Test no rejoin
* Test malformed m.call.member events
* Test event emitting
and also move some code to a more sensible place in the file
* Test getActiveFoci()
* Test event emitting (and also fix it)
* Test membership updating & pruning on join
* Test getOldestMembership()
* Test member event renewal
* Don't start the rtc manager until the client has synced
Then we can initialise from the state once it's completed.
* Fix type
* Remove listeners added in constructor
* Stop the client here too
* Stop the client here also also
* ARGH. Disable tests to work out which one is causing the exception
* Disable everything
* Re-jig to avoid setting listeners in the constructor
and re-enable tests
* No need to rename this anymore
* argh, remove the right listener
* Is it this test???
* Re-enable some tests
* Try mocking getRooms to return something valid
* Re-enable other tests
* Give up trying to get the tests to work sensibly and deal with getRooms() returning nothing
* Oops, don't enable the ones that were skipped before
* One more try at the sensible way
* Didn't work, go back to the hack way.
* Log when we manage to send the member event update
* Support `getOpenIdToken()` in embedded mode (#3676)
* Call `sendContentLoaded()` (#3677)
* Start MatrixRTC in embedded mode (#3679)
* Reschedule the membership event check
* Bump widget api version
* Add mock for sendContentLoaded()
* Embeded mode pre-requisites
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Embeded mode E2EE
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Encryption condition
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Revert "Embeded mode pre-requisites"
This reverts commit 8cd73702052609c995ad754e31f85d0da0be4aa9.
* Get back event type
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
fds
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Change embedded E2EE implementation
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* More log detail
* Fix tests
and also better assert because the tests were passing undefined which
was considered fine because we were only checking for null.
* Simplify updateCallMembershipEvent a bit
* Split up updateCallMembershipEvent some more
* Use `crypto.getRandomValues()`
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Rename to `membershipToUserAndDeviceId()`
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Better error
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Add log line
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Add comment
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Send call ID in enc events
(also a small refactor)
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Revert making `joinRoomSession()` async
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Make `client` `private` again
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Just use `toString()`
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix `callId` check
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix map
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix map compare
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix emitting
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Explicit logging
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Refactor
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Make `updateEncryptionKeyEvent()` public
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Only update keys based on others
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix call order
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Improve logging
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Avoid races
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Revert "Avoid races"
This reverts commit f65ed72d6e.
* Add try-catch
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Make `updateEncryptionKeyEvent()` private
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Handle indices and throttling
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Fix merge mistakes
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Mort post-merge fixes
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Split out key generation from key sending
And send all keys in a key event (changes the format of the key event)
rather than just the one we just generated.
* Remember and clear the timeout for the send key event
So we don't schedule more key updates if one is already pending.
Also don't update the last sent time when we didn't actually send the
keys.
* Make key event resends more robust
* Attempt to make tests pass
* crypto wasn't defined at all
* Hopefully get interface right
* Fix key format on the wire to base64
* Add comment
* More standard method order
* Rename encryptMedia
The js-sdk doesn't do media and therefore doesn't do media encryption
* Stop logging encryption keys now
* Use regular base64
It's not going in a URL, so no need
* Re-add base64url
randomstring was using it. Also give it a test.
* Add tests for randomstring
* Switch between either browser or node crypto
Let's see if this will work...
* Obviously crypto has already solved this
* Some tests for MatrixRTCSession key stuff
* Test keys object contents
* Change keys event format
To move away from m. keys
* Test key event retries
* Test onCallEncryption
* Test event sending & spam prevention
* Test event cancelation
* Test onCallEncryption called
* Some errors didn't have data
* Fix binary key comparison
& add log line
* Fix compare function with undefined values
* Remove more key logging
* Check content.keys is an array
* Check key index & key
* Better function name
* Tests too
---------
Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
Co-authored-by: David Baker <dave@matrix.org>
Co-authored-by: David Baker <dbkr@users.noreply.github.com>
https://github.com/matrix-org/matrix-js-sdk/pull/3756 changed
the membership update function to await on the next call, but this
meant it never returned and therefore never cleared
`updateCallMembershipRunning`. We therefore didn't send the updated
call member event when leaving, instead sending it whenever the next
poll interval arrived.
This changes it to only await if we are retrying, not if we're just
scheduling the next poll.
Fixes https://github.com/vector-im/element-call/issues/1763
* Don't use event.sender in CallMembership
I fell into another js-sdk trap: this is "only guaranteed to be set
for events that appear in a timeline" and not state events. It does
not say why. We only ever used it to get the sender user ID anyway,
so just use getSender().
* Fix test
Timo