I think this is very hypothetical, requiring a major version bump to drop an old connector version. What is the actual problem that would arise for users?
We have not required a major version to drop some old connectors that in the past, and if we want to continue "mono repo" style in Flink, I don't think we should have that as a constraint. It ties us to too much legacy code maintenance. Best, Stephan On Mon, Sep 16, 2019 at 1:43 AM Becket Qin <becket....@gmail.com> wrote: > Thanks for bringing this up, Stephan. > > I am +1 on dropping support for Kafka 0.8. It is a pretty old version and > I don't think there are many users on that version now. > > However, for Kafka 0.9, I think there are still quite some users on that > version. It might be better to keep it a little longer. > > Given that dropping connector support is a backwards incompatible change, > technically speaking we need to bump up the major version. I am not sure if > release 1.10 is a good version to remove the code. > Personally I prefer to mark Kafka 0.8 connector as deprecated and remove > the code base in Flink 2.x. > > Thanks, > > Jiangjie (Becket) Qin > > > On Wed, Sep 11, 2019 at 4:26 PM Wesley Peng <wes...@thepeng.eu> wrote: > >> >> >> on 2019/9/11 16:17, Stephan Ewen wrote: >> > We still maintain connectors for Kafka 0.8 and 0.9 in Flink. >> > I would suggest to drop those with Flink 1.10 and start supporting only >> > Kafka 0.10 onwards. >> > >> > Are there any concerns about this, or still a significant number of >> > users of these versions? >> >> But we still have a large scale of deployment kafka 0.9 in production. >> Do you mean the new coming flink won't support kafka 0.9? >> Though I understand for it, but just sounds sorry. :) >> >> regards. >> >
