-
Type: Task
-
Resolution: Done
-
Priority: Unknown
-
None
-
Affects Version/s: None
-
Component/s: AI/ML
-
None
-
Python Drivers
-
Not Needed
-
Context
We currently have been switching our CI to test against our own feature branches in CI. This is great during development, and also fine if the upstream maintainers plan to review and merge the work quickly. If there is a delay, and the PR becomes stale, we are not testing our integration with the one that users will be. We have already seen something like this coming out of our own bloggers experience.
config.yaml
- name: test-chatgpt-retrieval-plugin-rhel
display_name: ChatGPT Retrieval Plugin
expansions:
DIR: chatgpt-retrieval-plugin
REPO_NAME: chatgpt-retrieval-plugin
# TODO - Update CLONE_URL: [PYTHON-4291] [PYTHON-4129]
CLONE_URL: -b bugfix/broken-link-in-readme --single-branch https://github.com/caseyclements/chatgpt-retrieval-plugin.git
DATABASE: chatgpt_retrieval_plugin_test_db
run_on:
- rhel87-small
tasks:
- name: test-chatgpt-retrieval-plugin
Definition of done
I propose that we create NEW tasks when we are developing new features and when PRs are open. Not only will this continue to test upstream/main, it will also highlight to us that we need to change our config.yaml.