[79519] trunk/dports/www/neon

[Ryan Schmidt]

On Jun 17, 2011, at 02:55, Jeremy Huddleston wrote:

> On Jun 16, 2011, at 11:15 AM, Ryan Schmidt wrote:
> 
>> On Jun 16, 2011, at 11:18, jeremyhu at macports.org wrote:
>> 
>>> Revision: 79519
>>>        http://trac.macports.org/changeset/79519
>>> Author:   jeremyhu at macports.org
>>> Date:     2011-06-16 09:18:28 -0700 (Thu, 16 Jun 2011)
>>> Log Message:
>>> -----------
>>> neon: Build fix
>>> 
>>> neon uses gettext, so it needs an explicit -liconv.  It was previously getting
>>> it from one of its dependents, but that was likely moved into Libs.private.
>> 
>> You mean "an explicit -lintl"? That's what your patch adds. It builds fine for me on Snow Leopard without this patch though.
> 
> Yes, commit brain failure.
> 
> Where is it picking up the -lintl from?  It's not getting included here, and there is nothing in the autoconf logic that searches for it.  It was getting picked up indirectly through pkg-config for another package, so perhaps I have a newer version of some other package than you do.

As far as I can tell, -lintl comes from this bit of macros/neon.m4:

if test x${enable_nls} = xyes; then
  # presume that dgettext() is available if bindtextdomain() is...
  # checking for dgettext() itself is awkward because gcc has a 
  # builtin of that function, which confuses AC_CHECK_FUNCS et al.
  NE_SEARCH_LIBS(bindtextdomain, intl,,[enable_nls=no])
  NE_CHECK_FUNCS(bind_textdomain_codeset)
fi

Perhaps if you showed your config.log before your commit?