Include cause, whether on nightly or not

[curiousleo]

🐛 Set cause, whether we're on nightly or not.

Previously there were two cases with two sub-cases:

• if compiled with nightly:
  • if there is a panic message: set it as cause
  • if there is no panic message: set cause to "Unknown"
• if not compiled with nightly: set cause to "Error cause could not be determined by the application.", and also:
  • if there is a panic message: attach it straight to the expl string as "Cause: ."

That seems oddly complicated. This PR changes that to:

• if there is a panic message (no matter if we're on nightly or not): set it as cause
• if there is no panic message (no matter if we're on nightly or not): set cause to "Unknown"

Checklist

• tests pass
• tests and/or benchmarks are included
• documentation is changed or added

Context

Vaguely related to #55.

Semver Changes

This changes the format of the message users see. I think it's a bugfix, but if changing what users see is considered a breaking change, then this is a breaking change.

[curiousleo]

PTAL

[curiousleo]

PTAL

[Profpatsch]

pinging rust-cli WG members @codesections @Dylan-DPC @matthiasbeyer @XAMPPRocky

Our users cannot submit bug reports because of this bug, please merge and do point release, or we have to drop this package.

[matthiasbeyer]

I'd approve this change, though I'd do a minor version bump for this rather than a patchlevel bump.

@bjorn3 your review was requested, care to merge?

[bjorn3]

@matthiasbeyer I am not a member of @rust-cli.

[matthiasbeyer]

Oh,... why were you picked as reviewer by github then? I cannot merge, unfortunately.

[bjorn3]

I submitted a review, so github shows me in the list of reviewers.

[codesections]

I also don't have the correct permissions to merge this PR. It looks like @spacekookie would be able to, based on commit history.

[Dylan-DPC-zz]

Looks good to me. Merging it. Thanks @matthiasbeyer and special mention to @bjorn3 for showing up in the elite list of reviewers 😉

[Profpatsch]

Can I ask for a version bump nonetheless? #65