Hi @maintainers! This is an update on the upcoming Lilac upgrade. The new release is scheduled for June 9th – that’s Wednesday! We don’t have to produce a Tutor release on that same day, but I like to release Tutor as soon as possible after the new release is tagged. So I’d like to publish Tutor v12 before the end of the week.
There is still a little bit of work to do. Here’s the current status for the various components:
- Tutor (PR): I keep making small changes here and there as I discover new things during the upgrade process.
- tutor-discovery: See PR by @pcliu. A few minor adjustments need be made but nothing too difficult.
tutor-ecommerce (PR): this is a big one. The Ecommerce plugin will ship with both the payment and ecommerce micro frontends (MFE) apps enabled. Custom payment backends (other than Cybersource or Paypal) will no longer be supported. I will need to test whether these workflows created by Max Sokolski work with Tutor:
Test frontend-app-ecommerce native deployment · Issue #44 · openedx/build-test-release-wg · GitHub
Test frontend-app-payment native deployment · Issue #43 · openedx/build-test-release-wg · GitHub
My problem is that I don’t have valid Paypal or Cybersource credentials, and thus I’m not sure if I am even able to check that the new MFEs actually work. I could use some help here!
- tutor-mfe (PR): I am lagging behind with this PR. I still need to update the README to include proper docs. Also, I need to figure out how this is going to work on Kubernetes. I suspect I will simply ask people to build images locally and push images to a remote registry.
- tutor-minio (PR)
- tutor-notes (PR) Courtesy of @NeOneSoft
- tutor-xqueue (PR)
The tutor-figures plugin was deprecated some time ago, due to lack of upstream support.
In addition to these officially supported plugins, we should expect to get support of safe Python-evaluated inputs in Lilac, thanks to a Tutor plugin currently in development at eduNEXT: GitHub - eduNEXT/codejailservice This might not land in time for Lilac, but I think it’s perfectly fine since Python-evaluated inputs are no longer a security issue in Koa.