Re: Update for kbd-chooser patch (was: Re: r23706 - in trunk/packages/kbd-choose)

[Frans Pop <aragorn@tiscali.nl>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Thanks!
Great range of tests. Nice that you also tested serial installs.

AFAICT all tests lead to the expected results :-)
I've added some comments for some of the tests.

On Tuesday 09 November 2004 18:57, Joey Hess wrote:
> Results of my testing:
>
> ia64 at the console, usb keyboard, 2.6 kernel
>  choices were: USB keyboard, No keyboard
>  (Default was USB keyboard)
>  This seems wrong, why is PC-Style not an option?
>  However, after I configured it to use this keyboard, all the
>  keys on the keyboard seemed to work correctly, so maybe it was
>  right.

For the 2.6 kernel, the code in at-kbd.c checks for the presence of an AT 
connector (or maybe not for the connector itself, but for a kbd attached 
to it).
As I don't force the display of an AT keymap list for IA64 in usb-kbd.c, 
the result is correct with the current code.

That the keyboard works correctly is the result of the usb-mac-us keymap 
being the same as an AT keymap (unlike the usb-mac keymaps for e.g. FR 
and BE).

> vmware with usb disabled, 2.4 kernel
>  choices were: PC-Style, USB Keyboard, No keyboard
>  (Default was PC-Style)
>  FWIW, this system has no /proc/bus/usb/devices, so apparently
>  it's adding the USB keyboard choice even if there's no USB.

usbfs (which populates /proc/bus/usb) will be mounted (and unmounted 
afterwards) by kbd-chooser if it has not been mounted beforehand.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFBkUgBgm/Kwh6ICoQRAkFmAJ9ucS8FFOuMYY0tsmSBG1+Q6JIVhwCfb0MN
QHK1s7Gu0bjDtsjmXKigNBk=
=Mvn0
-----END PGP SIGNATURE-----