Prefer descriptive panic over unwrap

[oceanlewis]

See #23

[LegNeato]

I think it will be cleaner to just use expect()

[LegNeato]

Thanks for the PR! 🍻

[oceanlewis]

I was using the current implementation because I wanted whatever Display implementation there is for that error to do its thing. Does expect() automatically use that the Display impl for that type?

[srijs]

Another bit that might be nice to have is a custom error message. Currently we'll just wrap the raw env or parse error, which have fairly generic error messages, and it could be helpful to have a more descriptive message in these cases.

[oceanlewis]

Yeah, that would be nice. But at least you get more info than an expect("Failed to make a new Runtime"). That might be a separate issue than this PR?

[LegNeato]

Expect uses Debug for the underlying error rather than Display I believe:

https://doc.rust-lang.org/book/second-edition/ch09-02-recoverable-errors-with-result.html

[oceanlewis]

Okay, yeah - if you'd prefer that I'll change it.

[LegNeato]

Great, thanks! I think we will leave the more in-depth changes for another PR.