-
Notifications
You must be signed in to change notification settings - Fork 7
Wrong version assigned on updates #93
Copy link
Copy link
Open
Labels
kind/bugCategorizes issue or PR as related to a bug.Categorizes issue or PR as related to a bug.lifecycle/frozenIndicates that an issue or PR should not be auto-closed due to staleness.Indicates that an issue or PR should not be auto-closed due to staleness.priority/important-soonMust be staffed and worked on either currently, or very soon, ideally in time for the next release.Must be staffed and worked on either currently, or very soon, ideally in time for the next release.sig/stack-guidanceCategorizes an issue or PR as relevant to SIG Stack Guidance.Categorizes an issue or PR as relevant to SIG Stack Guidance.
Metadata
Metadata
Assignees
Labels
kind/bugCategorizes issue or PR as related to a bug.Categorizes issue or PR as related to a bug.lifecycle/frozenIndicates that an issue or PR should not be auto-closed due to staleness.Indicates that an issue or PR should not be auto-closed due to staleness.priority/important-soonMust be staffed and worked on either currently, or very soon, ideally in time for the next release.Must be staffed and worked on either currently, or very soon, ideally in time for the next release.sig/stack-guidanceCategorizes an issue or PR as relevant to SIG Stack Guidance.Categorizes an issue or PR as relevant to SIG Stack Guidance.
Type
Projects
Status
📋 Backlog
Describe the bug
See the following endpoint that lists data for versions. It states that elyra in version=3.5.1 (on UBI 8 Python 3.8) has a dependency on
click<8,>=7.1.1. However linked versions in the database are (based on API endpoint):I suspect this wrong version assigned in the database is caused by a release of
click==8.0.4. For some reason, revsolver assignsclick==8.0.4as a dependency ofelyra==3.5.1even though it does not respect the version specifier. I'm not sure if revsolver is really the issue maker, but the corresponding solver document that solved elyra==3.5.1 on UBI 8 Python 3.8 has noclick==8.0.4stated (and click in that version might not be even released at the time of solving).To Reproduce
I see this issue for the second time. Not sure how to reproduce it directly except for waiting for package releases and check timestamps for things that are happening in the system (or run revsolver locally against a local database dump without new releases). The issue might be caused by some incosystency (maybe environment markers?) that create this issue from time to time.