Re: RFC: Change to the way sh -x works?

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Wed, 15 Nov 2017 22:46:54 +0100
    From:        Edgar =?iso-8859-1?B?RnXf?= <ef%math.uni-bonn.de@localhost>
    Message-ID:  <20171115214654.GI73396%trav.math.uni-bonn.de@localhost>

  | I still prefer your original approach:

The new is better, you just haven't seen it yet.   It is more flexible,
more backward compatible, and just as easy to use - it just doesn't
automatically apply to uses of -x that already exist.

  | If I understood you correctly, the old behaviour can be emulated by
  | 	set -x 2>/dev/stderr

No, that would not work, with the original proposal there was no way
(unless a new option was implemented to disable the new) to keep the old
method.

2>/dev/stderr is a no-op in any case (it just says that fd 2 is directed to 
wherever fd 2 is directed, which is what /dev/stderr is, equiv to /dev/fd/2)

kre