On Sat, Jan 31, 2009 at 07:44:59PM +0200, Anton Zinoviev wrote: > I would like to revert r56948 unless there is something I don't see. As > far as I can remember this change increases the size of the precompiled > keymaps by many times. The precompiled keymaps will be used by udebs on > the boot media of d-i, so they have to take as little space as possible.
As you note, this is not in fact the case. > I think r56948 makes only the following improvement: Suppose a keyboard > layout doesn't define the behaviour of some of the keys on the keyboard. > For example the US layout doesn't define the behavour of one extra key > that exists on the Brazilian ABNT2 keyboards. Then suppose the user has > attached to the computer a Brazilian keyboard but selects US layout in > the installer. In this case before r56948 the extra key of the Brazilian > keyboard with US QWERTY layout generated some more or less random > symbols. After r56948 this key has undefined behaviour (for d-i this > means the key produces no symbols). The effect is similar when the user > has Japanese keyboard and non-Japanese layout. I think r56948 does > nothing when the keyboard is not Brazilian or Japanese (the US keyboards > do not have LSGT key that exists on Western European keyboards but > nevertheless the US layout defines behaviour for this key so r56948 does > nothing in this case). Here's the original bug report for this change: https://bugs.launchpad.net/ubuntu/+source/console-setup/+bug/74375 Obviously this is only a serious issue if you're using the keymap detection widget implemented in Ubuntu, but still - is it not an improvement? I think the logic is sound. -- Colin Watson [cjwat...@debian.org] -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
