[
https://issues.apache.org/jira/browse/CASSANDRA-18004?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17627819#comment-17627819
]
Josh McKenzie commented on CASSANDRA-18004:
-------------------------------------------
Are {{AsyncNotifier}} and {{AsyncChain}} _unique_ to accord or are they more
general purpose replacements for futures? If the latter, it seems a little odd
to have them live in Accord rather than live in Cassandra as a lot of that
shape of stuff came in with CEP-10 (CASSANDRA-16924); that said, it looks like
part of this effort was driven by the desire to break an accord -> cassandra
[dependency|https://github.com/bdeggleston/cassandra-accord/commit/f6c1319bac259292a5e82bf8a084468412bac974]
which makes sense to me and is worse than the other direction given accord's
goal of being C* independent. The straddle (some of these live in C*, some live
in Accord) stands out to me as a potential source of duplication and defects in
the future though.
This makes me wonder whether an Async / Futures library or subproject in the C*
build with these concurrency primitives that are going into Accord + the ones
that came into Cassandra with CEP-10 might make sense so we don't have either
the hard dependency between accord -> cassandra or the rather surprising
dependency of a general purpose Futures replacement we rely on that lives in
Accord. Would also allow us to build and test these structures independently of
Cassandra proper and rev them independently which could add some value as well.
> CEP-15: (C*/Accord) - remove futures
> ------------------------------------
>
> Key: CASSANDRA-18004
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18004
> Project: Cassandra
> Issue Type: Improvement
> Components: Accord
> Reporter: Blake Eggleston
> Assignee: Blake Eggleston
> Priority: Normal
>
> Remove futures in favor of async chain
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
