Dear fellow Nameko users,
I have a question regarding the plan forward for RabbitMQ.
I have been playing around with Nameko (2.12.0 and also 3.0.0rc8), and I noticed that in both cases, the automated tests rely on RabbitMQ version 3.6.6. This worried me a bit, as RabbitMQ 3.6 reached its end-of-life May 2018 (https://www.rabbitmq.com/versions.html). Currently, it is versions 3.7 and 3.8 that are supported, with the former’s EOL set to March 2020.
So I started playing around with different RabbitMQ versions locally, using the nameko-rabbitmq (https://github.com/nameko/nameko-rabbitmq) repo to build images for 3.8. Unfortunately, I found that both test suites of Nameko 2.12.0 and 3.0.0rc8 do not pass with RabbitMQ 3.8.
My questions are:
Has anyone else also seen this? I’m curious, specifically, if you know specifically which non-RabbitMQ 3.6 behavior that broke the tests. For the record, I found that simple applications with simple RPC calls can run just fine with RabbitMQ 3.7. However, Nameko’s test suite does cover much more ground than simple RPC services, and so I am worried that using Nameko with RabbitMQ >3.6 will lead to weird bugs in production.
What are the plans for ensuring that Nameko works well with current (and future) RabbitMQ releases?
While we are discussing versions, I also see that the latest test suite runs many different Python versions up to 3.6. Are there any plans to include Python 3.7 and 3.8 soon?
I’d be happy to help if needed.
With kind regards,