[Top][All Lists]

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

Re: [PATCH v2 0/4] Update gnulib and drop patches

From: Robbie Harwood
Subject: Re: [PATCH v2 0/4] Update gnulib and drop patches
Date: Thu, 27 Jan 2022 11:46:04 -0500

Glenn Washburn <> writes:

> Robbie Harwood <> wrote:
>> fix-width.patch has not merged upstream.  A version has been
>> proposed, but would need significant rewrite.  It doesn't seem worth
>> the effort to carry, and once this changeset merges to grub, it
>> becomes easier to increment our gnulib version if something like it
>> ever does merge there.
> Thanks Robbie for updating gnulib, this is something I've been wanting
> to get done. If I'm understanding you correctly, fix-width.patch is
> being dropped but there is not replacement for what it did.

All good questions.  To back up for a moment, my original plan had been
to wait longer and see if we could get a better version of that patch
accepted into gnulib.  However, we're accelerated from that now because
us using an old gnulib is holding up other efforts (and according to my
bug mail this morning, is needed to build with newer gcc/glibc).

> How was this patch functionally being used by GRUB? And what will the
> effects be of dropping it? Looking at the patch maybe one of the
> places its used is in GRUB's help command. If so, does that mean that
> line wrapping when printing help will be affected?

Not totally clear, and there's no commit information to help.  Mostly it
seems to change gnulib's width assumption to be multibyte-character
aware.  I think this could cause lines to not wrap early in locales that
are mutibyte-heavy, but I don't actually know, as I don't speak any
languages that meet that criteria.

gnulib upstream pointed out that this is only papering over the problem
though: it won't work for Chinese, for instance.  (It also won't work
for Thai, but there's nothing that anyone can do there.)

I could be wrong, but I'm actually not sure this affects grub at all.
We're not translated, which I think means that every character we put in
a help string will fit in a single byte under UTF-8.

Be well,

Attachment: signature.asc
Description: PGP signature

reply via email to

[Prev in Thread] Current Thread [Next in Thread]