Open epage opened 1 year ago
For test runs, a user would likely want to re-run the command and see the output, so probably not caching failures
For what it's worth, I cache failures too for my run
implementation. It's oftentimes expensive to go back to a commit and rebuild and rerun tests, so having the cached output is useful for me. I'm also guilty of running a test and then immediately forgetting the output and having to run it again.
I was wanting to avoid capturing stdin/stdout/stderr but to leave them as-is to avoid limitations around that (correctly detecting color, etc). When caching "success" I was just going to capture the status and report that.
Please complete the following tasks
Version
0.10
Use Case
Run faster by skipping runs where the result is the same as the past
Requirements
Possible Solutions
Cache successful command runs by commit and command flags in a json file in the
.git
directoryReport in the output that we are skipping due to prior success.
Disable with
--no-cache