Do you think this is a security risk and if not is it a bad UI decision?

[Nico Kadel-Garcia]

On Mon, May 6, 2013 at 8:34 AM, Bill Peck <bpeck at redhat.com> wrote:
> On 05/04/2013 06:22 PM, Dan Mashal wrote:
>>
>> On Sat, May 4, 2013 at 2:37 AM, Michael Scherer <misc at zarb.org> wrote:
>>>
>>> I can add to that that I have seen more than once people setting a
>>> password which was not the one they believed due to  :
>>> - keyboard layout ( ie, qwerty vs azerty in France )
>>> - small usage difference with Windows way, again on azerty keyboard
>>> ( people using capslock on french keyboard to type numbers while they
>>> should use shift, as capslock just type capital letter like À or É and
>>> not 0 or 2, and if you do not understand, just look on the web to
>>> compare how different it is from qwerty-based keyboard )
>>
>> The installer should detect the keyboard automatically. In fact you
>> can even tell it what type of keyboard you have on the first screen.
>
> On the screen where you can pick your keyboard, do we have a test area where
> the user can verify the keyboard layout?  Or maybe if you select a different
> keyboard it should automatically pop up a verify keyboard screen?

No matter how smart the kernel or anaconda, this can be nightmarish
when funneled through a virtualization console. The remapping between
the user's operating system, through the VMware or VNC or other
virtualization console access toolkit, can create..... some very odd
remappings.