[Nut-upsuser] Network UPS Tools 2.4.3 - compile on RHEL5

[Yury V. Zaytsev]

Hi!

On Thu, 2010-02-25 at 15:56 +0100, Arjen de Korte wrote:

> You didn't read "docs/configure.txt" nor did you run "./configure  
> --help". 

As surprising as it might seem, I did read ./configure --help, which
says this to be exact:

--with-gd-libs=LDFLAGS  linker flags for the gd library

As a matter of fact I didn't read "docs/configure.txt" of the existence
of which I was not even aware of, however, it does not tell anything
about appending "yes" the link flags if the link flags are not
explicitly specified.

It also does not explain a sudden change of the behavior of this option
from NUT 2.4.1 to 2.4.3, when a logical and intuitive functioning was
replaced with counter-intuitive and misleading at very least. In fact,
the error message is making it hard to recognize what is the actual
issue:

checking for gd library flags... yes

I would rather call it a regression that has to be fixed, which was
introduced by the replacement of this:

   AC_ARG_WITH(gd-libs,
   AC_HELP_STRING([--with-gd-libs=FLAGS], [linker flags for the gd library]),
   [<-->case "${withval}" in
<------>yes|no)
<------><------>;;
<------>*)
<------><------>LDFLAGS="${withval}"
<------><------>;;
<------>esac],
   )

by this:

<------>AC_ARG_WITH(gd-libs, [
<------><------>AC_HELP_STRING([--with-gd-libs=LDFLAGS], [linker flags for the gd library])
<------>], [LDFLAGS="${withval}" LIBS=""], [])

> Please read the documentation before asking for help.

I am grateful for the pointer that solved my problem, but I wonder why
your reply couldn't be just a little bit friendlier, considering the
fact that it's not an easy job for the packagers to follow random
illogical changes in upstream behavior and this is exactly the case.

This option was added by previous Redhat maintainer and I didn't even
notice it, and when I did, I assumed that it was there for a reason.
 
-- 
Sincerely yours,
Yury V. Zaytsev