[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: patch


From: Petter Reinholdtsen <pere@hungry.com>
Subject: Re: patch (Re: i18n of cdebconf)
Date: 31 Aug 2003 22:02:43 +0200

> I read your patch, and have a few comments.  It was very small.  That
> is good.
> If you protect
> the code like this, and keep the library code out of the cdebconf
> source, feel free to commit the change. :)

Thank you for checking my patch.  I will revise it and will
(maybe) commit it.  Fortunately, libtextwrap1 and libtextwrap-dev
packages are installed into Sid recently.

> Also, I want to make sure all parts of the installer use the same
> shared libtextwrap, to limit the disk image size.

Ok.  My patch will make conventional text-wrapper surrounded by
#ifndef HAVE_LIBTEXTWRAP and #endif.

> Regarding the library itself, I am not convinced that this charset
> normalization is a good idea:
>   Since the return value of nl_langinfo() may differ depending on
>   systems (such as "EUC-JP" vs "eucJP"), the encoding names are
>   normalized (all-uppercase, eliminate non-alphanumerics).
> It is better to find and fix all systems not using a standard charmap
> name.  But this will be hidden in the library, so cdebconf do not
> really need to worry about it.

Such a normalization is not needed if the platform is limited to
GNU/Linux.  However, I tested FreeBSD and it had different naming
way for locales (for example, ISO8859-1 in FreeBSD vs ISO-8859-1
in GNU libc).  Moreover, there are many proprietary UNIX systems.

Tomohiro KUBOTA <kubota@debian.org>

Reply to: