Subject: gnupg 1.4 and libiconv (was Re: gnupg 1.4)
To: None <tech-pkg@netbsd.org>
From: Christopher W. Richardson <cwr@nexthop.com>
List: tech-userlevel
Date: 01/06/2005 11:49:03
Thomas Klausner <wiz@NetBSD.org> writes:

> On Wed, Jan 05, 2005 at 11:42:27AM -0500, Christopher W. Richardson wrote:
> > I'm attempting to upgrade to gnupg 1.4 from pkgsrc.  Compilation
> > "appears" to be fine, but the make never finishes due to failures
> > in gnupg-1.4.0/checks. All of the checks fail thusly:
> > 
> > Making all in checks
> > gmake[2]: Entering directory `/usr/pkgsrc/security/gnupg/work.i386/gnupg-1.4.0/checks'
> > ./gpg_dearmor > ./pubring.pkr < ./pubring.pkr.asc
> > gpg: conversion from `utf-8' to `646' not available
> 
> You're probably compiling on 1.6.x against the pkgsrc libiconv.

You are correct. I'm running

NetBSD achilles 1.6.2_STABLE NetBSD 1.6.2_STABLE (ACHILLES) #7: Mon Jan  3 11:05:50 EST 2005 cwr@achilles:/usr/src/sys/arch/i386/compile/ACHILLES i386

and libiconv-1.9.2nb1 from pkgsrc.

> Adding the following line to /usr/pkg/lib/charset.alias should
> fix the problem:
> utf-8 ASCII
> 
> However, I'm not completely sure if it's correct.  Someone here
> know enough about (lib)iconv to confirm it is correct?

Thank you very much for finding a starting point to look into the
mystery.  My /usr/pkg/lib/charset.alias is this:

# This file contains a table of character encoding aliases,
# suitable for operating system 'netbsdelf'.
# It was automatically generated from config.charset.
# Packages using this file:  gettext
646 ASCII
ISO8859-1 ISO-8859-1
ISO8859-2 ISO-8859-2
ISO8859-4 ISO-8859-4
ISO8859-5 ISO-8859-5
ISO8859-15 ISO-8859-15
eucCN GB2312
eucJP EUC-JP
eucKR EUC-KR
eucTW EUC-TW
BIG5 BIG5
SJIS SHIFT_JIS

Unfortunately, adding your "utf-8 ASCII" suggestion, either
before or after the "646 ASCII" line, or replacing the 646 with
utf-8 does not solve the problem.  I will continue to dig into
this now that I have a starting point, but if anyone knows
anything about libiconv, hints would be much appreciated.

Thanks,
Chris