Sorry if there is an obvious answer to this question that I'm missing,
but what is the advantage of using translation in set 3?
I totally agree that translated set 2 is the way to get 99% of
keyboards working perfectly, and that the reason we use translation
here, is because although untranslated set 2 is simpler, some laptops
don't support this properly, and some that do have problems with BIOS
interpretation of the codes, etc.
However, surely setups that support set 3, will support it equally
well with and without translation? Here, I don't see the advantage of
enabling translation.
Why not simplify the whole problem, and either have:
* translated set 2 with workarounds for all known strange keyboards
* untranslated set 3
John.
On Mon, Sep 22, 2003 at 01:42:55PM +0100, John Bradford wrote:
> Sorry if there is an obvious answer to this question that I'm missing,
> but what is the advantage of using translation in set 3?
>
> I totally agree that translated set 2 is the way to get 99% of
> keyboards working perfectly, and that the reason we use translation
> here, is because although untranslated set 2 is simpler, some laptops
> don't support this properly, and some that do have problems with BIOS
> interpretation of the codes, etc.
>
> However, surely setups that support set 3, will support it equally
> well with and without translation? Here, I don't see the advantage of
> enabling translation.
>
> Why not simplify the whole problem, and either have:
>
> * translated set 2 with workarounds for all known strange keyboards
> * untranslated set 3
Because it doesn't simplify the driver. :) The translated+set3 combo,
although doesn't make much sense, just works ...
--
Vojtech Pavlik
SuSE Labs, SuSE CR
On Mon, Sep 22, 2003 at 01:42:55PM +0100, John Bradford wrote:
> Why not simplify the whole problem, and either have:
>
> * translated set 2 with workarounds for all known strange keyboards
> * untranslated set 3
Yes, you need untranslated set 3, everybody else on i386 needs
translated set 2.
(Other architectures do not do any translation and probably need
untranslated set 2.)
(Very rarely, one sees untranslated set 1 on i386.)
Let us wait and see what Vojtech creates.
On Mon, Sep 22, 2003 at 10:25:27PM +0200, Andries Brouwer wrote:
> > Why not simplify the whole problem, and either have:
> >
> > * translated set 2 with workarounds for all known strange keyboards
> > * untranslated set 3
>
> Yes, you need untranslated set 3, everybody else on i386 needs
> translated set 2.
>
> (Other architectures do not do any translation and probably need
> untranslated set 2.)
>
> (Very rarely, one sees untranslated set 1 on i386.)
>
> Let us wait and see what Vojtech creates.
How about this? Warning: COMPLETELY UNTESTED. But it should give you an
idea where I'm heading to.
--
Vojtech Pavlik
SuSE Labs, SuSE CR