To:         J3                                      J3/16-132r1
From:       Van Snyder & Dan Nagle
Subject:    Descriptions of BESSEL_JN and BESSEL_YN are wrong
Reference:  16-007
Date: 2016 February 10

1. Introduction
===============

The descriptions of the result characteristics of BESSEL_JN and
BESSEL_YN are incorrect for the elemental cases.

Does this need to be corrected by an interp?

The shape of the result of an elemental function is scalar.  The shape
of the result of applying an elemental function elementally is
adequately described in 12.8.2.  Clause 13 does not describe the
result of applying any other elemental intrinsic function.  If we
must specify the shape in the elemental cases for BESSEL_JN and
BESSEL_YN, it's not the same as X.  It's the same as N if N is
an array, else the same as X.

2. Edits
========

[348:3-4] 13.8.33p3 BESSEL_JN insert "shall be"
before "an integer" twice

{make argument descriptions consistently}

[348:7-8 13.8.33p4(i)] Delete "Case (i) ... Case(ii):"
and combine with the previous line.

{state result characteristics correctly for the elemental case}

The new Results Characteristics paragraph is
"Same type and kind as X.  The result of BESSEL_JN(N1,N2,X)
is a rank-one array with extent MAX(N2-N1+1,0)."

[348:2-3] 13.8.36p3 BESSEL_YN insert "shall be"
before "an integer" twice

{make argument descriptions consistently}

[349:7-8 13.8.36p4(i)] Delete "Case (i) ... Case(ii):"
and combine with the previous line.

{state result characteristics correctly for the elemental case}

The new Results Characteristics paragraph is
"Same type and kind as X.  The result of BESSEL_YN(N1,N2,X)
is a rank-one array with extent MAX(N2-N1+1,0)."