* 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
* Fix issue with sentinels being incorrect on m.room.member events
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Iterate
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Simplify change
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Add test
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
---------
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* 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>
* Don't retry on 4xx responses
I'm not sure why this was limited to a small set of 4xx responses.
Nominally, no 4xx request should be retried (in fact the comment
below says this, but then the code didn't quite match it).
This was causing key backup requests to be retried even when the
server responded 404 because the backup in question had been deleted,
meaning the client would retry uselessly and it would take longer for
the client to prompt the user for action.
* Exclude 429s
* Avoid key prompts when resetting crypto
Attempting to get the backup key out of secret storage can cause
the user to be prompted for their key, which is not helpful if this
is being done as part of a reset. This check was redundant anyway
and we can just overwrite the key with the same value.
Also fix docs and remove check for active backup.
* Fix doc
* Save the key backup key to secret storage
When setting up secret storage, if we have a key backup key in cache
(like we do for the cross signing secrets).
* Add test
* Get the key directly from the olmMachine
saves converting it needlessly into a buffer to turn it back into
a base64 string
* Overwrite backup keyin storage if different
* Fix test
* Add integ test
* Test failure case for sonar
* Unused import
* Missed return
* Also check active backup version
* 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>
David Baker