Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
sysctl: Rename ignoreerrors to ignore_unknown_keys #49253
This PR renames ignoreerrors to ignore_errors and keeps backward compatibility with ignorerrors.
It's basically to be consistent with all other modules.
The sysctl module is the only one where the attribute is called ignoreerrors and not ignore_errors.
With this PR ignore_errors is allowed, too. I found it a bit unintuitive because it is not consistent with all other modules where the parameter is called ignore_errors with an underscore. The alias is to ensure backward compatibility because this cosmetically change should not break any playbooks out there.
Documentation was updated and 2 tests to test both, ignoreerrors and ignore_errors were added.
@bcoca I totally agree here. It's a bit confusing especially when you combine both or use both in the same playbooks. To be honest, it even took me a moment to understand the different effects of ignoreerrors and ignore_errors.
So I took another look at the other modules and it's really just the keyword that it's referenced there.
I have mixed feelings about this PR and am not sure if consistency is worth the risk for confusion here. I let this up to you, guys. I also don't mind to change anything here.
(Will take a look at the failing tests for Fedora, CentOS, etc. tomorrow. Already 3 a.m. here.)
I thought a bit more about this and, if at all, it probably just makes sense to choose another name for the parameter (i.e. ignore_unknown_keys). ignore_errors on 2 different levels is something that should be avoided.
The reason why I changed my opinion a bit is this from the previous comment: "So I took another look at the other modules and it's really just the keyword that it's referenced there."
If there is no further input or request from anyone else with an opinion on this toic, I'll close this PR.
@bcoca what do you think about the option to rename it to "ignore_unknown_keys"? I think that this is a small improvement and not as ambiguous as