[SR-8143] Should RNG.next(upperBound: 0) be a preconditionFailure?

[stephentyrone]

Previous ID	SR-8143
Radar	rdar://problem/41735302
Original Reporter	@stephentyrone
Type	Bug
Status	Resolved
Resolution	Done

Additional Detail from JIRA

Votes	0
Component/s	Standard Library
Labels	Bug
Assignee	@stephentyrone
Priority	Medium

md5: aa7632d8600c5ad6ba8cbd93aa8f3d7c

Issue Description:

Presently we have the following:

guard upperBound != 0 else { return 0 }

This seems like a pretty glaring violation of the documented behavior: "Returns a random value that is less than the given upper bound."

[stephentyrone]

It's entirely possible that I simply missed the discussion where this was decided. CC @natecook1000 @airspeedswift

[lorentey]

I agree, we should clearly trap here.

[stephentyrone]

I'll fix this on master and we should consider pulling it into 4.2.

[stephentyrone]

#17627

[stephentyrone]

and #17631 for 4.2

[airspeedswift]

@swift-ci create