Date: Mon, 29 Jul 2019 01:02:37 -0700
From: L A Walsh <b...@tlinx.org>
Message-ID: <5d3ea81d.4020...@tlinx.org>
| I didn't see this come back from the list and it was sent
| 45 minutes ago (vs. other emails of mine that have come
| back in under a minute).
| Did anyone else see it?
Yes. The original appeared (before this repeat). Avoid
being impatient, all kinds of things can delay e-mail (a once
an hour retry is not uncommon - so 45 mins is certainly not
long enough to wait before becoming concerned - allow at least
a day).
| FWIW, the below is from the publicly posted change notes
| for the 5.2.3 patch.
I have no idea what patch you're talking about, as it has
nothing to do with any system I run. I also don't understand
your concern.
I agree that simply "shove in 'set -e'" is often a very poor
idea, -e is a very difficult option to use correctly, and almost
always the lazy way to solve a problem.
Yet, it can be used if the script is analysed correctly
(-e was designed and intended as assistance for make)
and nothing in your message suggests to me that in whatever
script it is that the patch you are concerned about is
using -e in a way that is incorrect or liable to produce
undesired results. (I also don't know that it is
being used properly).
Whatever this is however, a possible incorrect usage in some
build system's scripts is most certainly not a bug in bash,
and does not warrant a message here. It has nothing to do
with bash at all that I can tell, unless the script in question
is so badly written that only bash can interpret it.
kre
