eng: loosen rubocop restrictions ECOMM-7882

[tshedor]

Inspired by this discussion

[8FAX]

I think this is set to false by default right, are we just adding it for clarity that more then 1 expect per it block is wrong?

[tshedor]

It's enabled by default

[8FAX]

So by setting this false we are allowing multiple expects per it block?

[tshedor]

Yes

[tmyrick]

I'm not in favor of this one but not a huge deal

[tmyrick]

rspec doesn't do a good job at letting you know which expectation failed when there are multiple

[8FAX]

I agree with Travis here, I think if we're going to loosen restrictions on application logic, we should still try to maintain the current test restrictions until we have a better idea of how this change actually will affect the code output.

[tshedor]

@tmyrick @8FAX This is genuinely my least favorite lint and the one I disable the most. But if you're both on board with keeping it I'll remove.

[8FAX]

Thanks for removing this, I think we may get to a point where this can be removed, but for now I think its helpful. I am more then happy to talk about this later on if we want to move to remove this.

[evan-waters]

I somewhat agree with the intended purpose of this lint. It prevents you from having hidden failures in a single test, but I do not necessarily agree with it as a style rule. I favor using :aggregate_failures on my test block to get around this lint. The rule allows this without having to ignore.

[danhealy]

I'd be disappointed to reduce our code quality targets without data to justify it.

I think if we wanted to avoid bikeshedding, we should consider using https://github.com/standardrb/standard