Revert "Pipenv (#4494)"#4538
Conversation
This reverts commit 4c0a60d. Signed-off-by: Eric Stroczynski <ericstroczynski@gmail.com>
|
/cc @reinvantveer |
|
@estroz: GitHub didn't allow me to request PR reviews from the following users: reinvantveer. Note that only operator-framework members and repo collaborators can review this PR, and authors cannot review their own PRs. DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. |
|
Yikes! |
|
@estroz This github actions workflow may be of interest to you: https://github.com/joelanford/helm-operator/blob/main/.github/workflows/deploy.yml It runs the goreleaser and docker builds on all PRs, but only publishes on tags. |
|
I can see pipenv retrying a couple of times on some hashes, but so far I can see builds failing only on missing gcc. So perhaps if we install and de-install gcc as a build dependency for the python dependency phase, things will be OK? |
retry of operator-framework#4538 and operator-framework#4494 Signed-off-by: reinvantveer <reinvantveer@gmail.com>
It doesn't appear to be the only issue. After adding gcc (along with |
|
OK I' ll give it a try |
|
#4543 is where I intend to get things going again, if possible |
retry of operator-framework#4538 and operator-framework#4494 Signed-off-by: reinvantveer <reinvantveer@gmail.com>
retry of operator-framework#4538 and operator-framework#4494 Signed-off-by: reinvantveer <reinvantveer@gmail.com>
retry of operator-framework#4538 and operator-framework#4494 Signed-off-by: reinvantveer <reinvantveer@gmail.com>
retry of operator-framework#4538 and operator-framework#4494 Signed-off-by: reinvantveer <reinvantveer@gmail.com>
This reverts commit 4c0a60d.
#4494 was merged prematurely. Builds for s390x, ppc64le, arm64 are broken due to gcc (and other deps) not being installed, but during some testing I also discovered that the Pipfile.lock's hashes don't resolve on s390x builds (at least) and I can't debug the issue. #4494 should be reverted until builds on all platforms can be tested using Github Actions for each PR, since we're moving away from Travis completely soon. This is something I'm working on right now using
docker buildx.