Re: [PATCH] Avoid possible false negatives in cond46.test.

[Stefano Lattarini]

At Sunday 11 April 2010, Ralf Wildenhues <address@hidden> 
wrote:
> > How'd you suggest to shorten the log entry above, while keeping
> > it meaningful?
> 
> By cutting the part which I did not quote.  :-)
Oops, I missed that.
> 
> Being verbose is fine for things that are unobvious, but you may
>  assume that developers know autoconf.info.
Quite right.  However, as you said when discussing my proposed patch 
enabling `set -e' in all test scripts, the behaviour of the `errexit' 
shell flag can vary wildly among different shells in unusual situations, 
and is generally quite bug-prone when control structures or complex 
commands are involved.  The more detailed analysis contained in the 
git commit message tried to show (by static code inspection only) that 
no bug should have cropped up in the test case in question.  However, 
if you deem that such an analysis is not required, I'll be happy to 
oblige in future similar patches.  Really happy indeed ;-)

Regards,
    Stefano