[
https://issues.apache.org/jira/browse/SAMZA-18?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14065477#comment-14065477
]
Chris Riccomini commented on SAMZA-18:
--------------------------------------
David, I saw you grabbed this ticket. I am concerned with working on this
ticket right now. If we take for granted that a multi-lang C library is useful,
there is a lot of work to be done here. Our existing JVM implementation is
still evolving at a really high rate. I believe that this is because we're
still learning about the best way to implement certain features, and
discovering new use cases that we hadn't though of before. Until we get the JVM
implementation settled down, and are happy with its implementation, I don't
think we should start on a C implementation.
A related and more approachable feature is SAMZA-184, if you're interested.
What do you (and others) think?
> Create single multi-lang library in C
> -------------------------------------
>
> Key: SAMZA-18
> URL: https://issues.apache.org/jira/browse/SAMZA-18
> Project: Samza
> Issue Type: Wish
> Components: container
> Affects Versions: 0.6.0
> Reporter: Travis Brady
> Assignee: David Chen
> Labels: project
>
> I suggest that a single canonical C library be written to implement most of
> the multi-lang support.
> A library of this sort would allow people using Python, Ruby, OCaml, Haskell,
> Node, Lua, etc etc to simply bind to the C library and expose all Samza
> functionality without having to implement it directly.
> As an example, there are a handful of Kafka clients for Python and they are
> spotty in terms of their support for the full feature set.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
