New Issue: Should bounded fillRandom() calls change to exclusive upper bounds?

20165, "bradcray", "Should bounded fillRandom() calls change to exclusive upper bounds?", "2022-07-06T17:46:38Z"

In this comment chain on Arkouda we realized that both Python and PCG use an exclusive upper bound, where Chapel's fillRandom() was defined as being inclusive. It seems as though changing it to be exclusive would be more consistent / less surprising with what users might expect. This issue proposes that we change to exclusive while we have separate min/max values.

[It was also suggested that maybe we should take a range to indicate the bounds which would have the advantages of (a) making the inclusive vs. exclusive nature more explicit (i.e., 0..n vs. 0..<n), similar to how we're changing channel creation to take its region as a range... But a challenge in this context is that we don't currently support ranges of floating point values]