Michael Barrientos

Fact of the matter is that it's NOT being caught in plan phase, otherwise I wouldn't have written this. See #472 for how to reproduce. Verified a problem in TF...

Did some more investigation into @mcuadros assertion that the #464 should be enough to catch the problem in the plan phase. I found that the error I'm catching here is...

Please see #516 for alternate fix. I need feedback to know which approach is the better one, given that #516 breaks some tests but in deprecated code. The underlying bug...

This may or may not be the same bug, but here's a way to get into a similar state, which I ran into inadvertently when trying to fix repository names....

So the only failure I could reproduce with a real bug is the one I described in my previous comment. The key lines of the debug output are ``` 2020/05/30...

As for the originally described error from the issue, I could not reproduce it. I tried the following to try to simulate a probe failing, using TF Helm provider 1.2.1:...

I reproduced my first use case above in 1.2.1 just yesterday. Do not close. Please re-add bug label.

While we're here, may also want to future proof for other cases we want to add things to a module, such as Terraformm 0.13's depends_on, or possibly for_each/count

Since it's such an easy check for the existence of the single line now, it would be nice if there was a feature to add this line to .ssh/config.

One possible solution is switching the underlying Auth0 library to use [Jettison](https://github.com/wl2L/jettison) for its JSON encoding, and use the omitnil tag instead of omitempty (which would omit nils, but keep...