• README: Explain how to run tests
• setup.py: Use a shell script to ease the kafka development setup
• Makefile.local: Add a target to clean up cache artifacts

So, if we keep the setup.py snippet to install kafka, I think it should be a completely separate subcommand of setup.py, rather than overriding the default behavior of the develop subcommand, which has a somewhat standard meaning of "install the current module for a development setup".

So, if we keep the setup.py snippet to install kafka, I think it should be a completely separate subcommand of setup.py, rather than overriding the default behavior of the develop subcommand, which has a somewhat standard meaning of "install the current module for a development setup".

Build is green
See https://jenkins.softwareheritage.org/job/DJNL/job/tox/43/ for more details.

Remove unused import

Build has FAILED

Rebase on latest master

I think i'm done for the main part of the tests.

Build has FAILED

Build has FAILED

Make test work for major object types

Build has FAILED

Rebase on latest master (and removes the cli adaptation part, D1249)

Build is green
See https://jenkins.softwareheritage.org/job/DJNL/job/tox/37/ for more details.

Adapt according to review

Thanks for the remarks.

Accepted, but let's not forget to add tests for the CLI

Build has FAILED

Will remove the 'develop' step later

In D1239#26394, @olasd wrote:

Thanks for looking into pytest-kafka!

Before you go too much further with this approach, I'll just say that it makes me very uncomfortable...

Thanks for looking into pytest-kafka!

Ensure this now rather than later ;)

Build is green
See https://jenkins.softwareheritage.org/job/DJNL/job/tox/28/ for more details.

heads up btw: It's deployed.

D918 closes this.

It's been done for the schema at that time so closing it.