Re: Bug#577253: initramfs-tools: Keyboard is not active / detected anymore, so LVM boot passphrase entry is broken

[Colin Darie <colindarie@gmail.com>]

No I think it's great to use the *real* keymap, but this change could be confusing because when users are entering the passphrase, they have the habit to use a qwerty keymap, and as they have no way to see that the keymap has changed, they might not be able to boot their system.

I'd suggest to add an entry into the NEWS file and in the release notes.

--------
Colin Darie

On Mon, Apr 12, 2010 at 4:22 PM, maximilian attems <max@stro.at> wrote:

On Mon, Apr 12, 2010 at 12:40:02PM +0200, Colin Darie wrote:
>
> Furthermore, before my keymap is this prompt was for a qwerty
> keyboard, and now this is the right keymap for my keyboard (a french
> one). It's great, but I think it might be useful to notify users for
> this change for the local keymap, after month / years with in qwerty
> for the LVM prompt.
>
> However, I don't know if this change is related to initramfs-tools
> package or the last cryptsetup update (2:1.1.0-2.1), because my
> current initrd image is the first one after the cryptsetup update few
> days ago.

hmm I'm really not an input expert, rereading the patch
0830e97339add137d56d13cdbe07282f3cdba8f6, that I merged for
considering /etc/console-setup/cached.kmap.gz seems wrong as it
now prefers /etc/console-setup/cached.kmap.gz over
/etc/console/boottime.kmap.gz
http://git.debian.org/?p=kernel/initramfs-tools.git;a=commitdiff;h=0830e97339add137d56d13cdbe07282f3cdba8f6;hp=323005e4b8468ab9f7151d883d3df7b75f130911

you could revert that in /usr/share/initramfs-tools/hooks/keymap
thanks for your feedback

any input from debian boot listed as console-setup maintainer on their
meaning?