Re: SegFault in stringsubst

[Peter Stephenson <p.w.stephenson@xxxxxxxxxxxx>]

On Wednesday, 16 April 2014, Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>
wrote:
> On Apr 16,  4:19pm, Andrew Waldron wrote:
> }
> } On Tue, 15 Apr 2014, Bart Schaefer wrote:
> }
> } > The exec.c change doesn't change anything other than verbosity about
some
> } > error conditions, correct?
> }
> } Correct, if you take the change in stringsubst then there is no
segfault.
> } You just end up with empty substitutions.
>
> Any second opinions on whether we want the increased verbosity?  Peter?

I think it's certainly useful in debug mode, and there's definitely an
argument for
always reporting it as the alternative is probably rather obscure. The only
likely
negative effect I can think of might be of the caller also reported an
error.

I'm likely to be quiet for a week.

pws