-
- [ ] Version HRMP channels
- Without versioning, the two issues below are much harder to solve.
- [ ] Double encode XCMs in HRMP channels
- for ensuring that one invalid message will not bre…
-
This would of course potentially slow everything down and could lead to nasty deadlocks if not implemented properly.
-
Currently we run user callbacks like On handlers, OnClosed, and possibly invocation responses/streaming. We want to run these on a dedicated scheduler that is used only for user code. We also want to …
-
Hi,
I have library https://github.com/cashapp/sqldelight/blob/master/extensions/coroutines-extensions/src/commonMain/kotlin/com/squareup/sqldelight/runtime/coroutines/FlowExtensions.kt#L36
which u…
-
Afaik it can happen that if the client executing a partition specific operation, that the invocation of this operation receives backpressure while running on the operation thread. This can cause big …
-
Given that the original desire was for a backpressure-aware protocol (not a protocol that avoids backpressure), I'd like to see a rationale of tradeoffs of each approach.
* Push-based with backpres…
raulk updated
4 years ago
-
At some point we should probably add backpressure so that very large writes do not use up all the memory - for now lets not worry about it unless you are feeling up to the challenge.
_Originally po…
-
Either we need to add backpressure to SubstraitSinkConsumer or (possibly easier) change SubstraitExecutor to use a SinkNode instead of a ConsumingSinkNode.
**Reporter**: [Weston Pace](https://issues.…
-
The new support to not backpressure for commit replies should be promoted to official API or deprecated and be removed again.
This is a discussion issue for reference from the source code.
ennru updated
5 years ago
-