Hello folks, I am curious about the current state ...
# troubleshooting
a
Hello folks, I am curious about the current state of development for Stateful Functions. The repo seems to be quite inactive and the last release seems to be a year old and the docker image is based on Flink 1.14.x. Has the community moved to a new solution that I need to be aware of? Is anyone actively using this project? Or is the lack of new development is because this is a thin layer on top of Flink and doesn't need many new features? Thank you so much!!
plusone 1
s
You’ll probably get more responses in the dev mailing list.
d
I have the same question. Interested to know if the maintainers of the repo have some thoughts/inputs here. Stateful functions seems to be falling behind the Flink release schedule, from looking at the two: https://github.com/apache/flink/tags and https://github.com/apache/flink-statefun/tags. On a slightly related note, here’s one of my other questions: https://apache-flink.slack.com/archives/C03G7LJTS2G/p1674240128177009.
➕ 1
a
I would really appreciate some pointers on this! I think the remote functions model is really what we need but it is concerning to see the inactivity on the repo. Do people still use statefun in production or did they move to something else for remote functions. Or maybe there is an inherent problem with using remote functions? Or is it just a matter of resources that the repo hasn't been active recently? I also asked this on the dev mailing list but hasn't received a response there either. Thanks!
plusone 3
a
Tagging @Tzu-Li (Gordon) Tai since you’re a frequent committer to statefun… Could you comment on the state of stateful functions?
s
+1, we heavily use stateful functions in production and it would be very problematic if this became deprecated. I also have a PR plus an issue open that hasn’t been looked at in forever.
g
The original developers of the project are not working on this anymore . Most of them are not working on Flink in general . I think if this is important for you, it's worth considering stepping up and starting to contribute more to it to keep it going forward.
🙏 1
👀 1
👍 2
I know this would be a big investment for all involved parties, but with collaboration it should be doable :)
m
+1 for what Gyula says.
a
This was the clarification I needed, thank you!
plusone 1
v
So basically the project is not maintained anymore ?
d
That’s what I’m reading between the lines here! Thanks for the clarification on this.
☹️ 1
a
Thank you for the clarifications @Gyula Fóra @Martijn Visser ! :)
t
I've been working on a statefun system for my company that I'm nearing production release on. I need to go through the process of becoming a contributor as I've added pubsub and elasticsearch ingress/egress and updated protoc I'd love to get into a release at some point. I absolutely love the framework.
v
We also have created a kafka + avro ingress / egress that might help other. Indeed it would be a shame if this framework died is as it simplifies a lot (in our case) data streams implementations.
g
This is open source, nothing dies that has contributors/companies willing to make an effort for. I am sure the community would be happy to mentor new statefun contributors, but users will have to step up because they can no longer expect someone else to add new features or develop this further 🙂 I encourage every one of you who are currently using statefun and want to keep doing it to pick up some easy tasks and make some improvements. I understand that this may require more time assigned for this from your employer but if you are invested in the product it should not be a problem. It will also greatly benefit you as a user over the long term if you take better ownership.
👍 5
➕ 2
s
@Gyula Fóra thanks for the insight! I think it would be worthwhile to announce this more broadly on a mailing list since this info is quite buried in a Slack thread. It’s quite a coincidence that I came across this thread and it this might be a game changer for us.