belliottsmith commented on code in PR #10:
URL: https://github.com/apache/cassandra-accord/pull/10#discussion_r990689184
##########
accord-core/src/main/java/accord/messages/Apply.java:
##########
@@ -54,15 +63,59 @@ public Apply(Node.Id to, Topologies topologies, TxnId
txnId, Txn txn, Key homeKe
this.result = result;
}
+ @VisibleForImplementation
+ public Apply(Keys scope, long waitForEpoch, TxnId txnId, Txn txn, Key
homeKey, Timestamp executeAt, Deps deps, Writes writes, Result result)
+ {
+ super(scope, waitForEpoch);
+ this.txnId = txnId;
+ this.txn = txn;
+ this.homeKey = homeKey;
+ this.executeAt = executeAt;
+ this.deps = deps;
+ this.writes = writes;
+ this.result = result;
+ }
+
+ static Future<Void> waitAndReduce(Future<Void> left, Future<Void> right)
Review Comment:
Honestly, the more I think about it, the more I am simply -1 the use of
`Future` where it isn't necessary. To some extent there has to be some design
consistency, and by baking `Future` in here we are guaranteeing their
widespread use forever more. I disagree with foisting this performance
trade-off on the hot-path, even if we reduce its cost.
The efficient way to receive a result is to declare the receiver upfront, as
there is no concurrency. There's no getting around this fundamental design flaw
with `Future`, at least not without additional memory costs.
If we have to, we can have a project-level discussion around this. But, to
the extent possible, I may simply invoke primary author privilege here.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
