[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [virt-tools-list] [libosinfo 1/2] install-config: Document expected kbd layout format



On Fri, Nov 30, 2012 at 05:38:16PM +0200, Zeeshan Ali (Khattak) wrote:
> On Fri, Nov 30, 2012 at 5:03 PM, Christophe Fergeau <cfergeau redhat com> wrote:
> > On Fri, Nov 30, 2012 at 04:30:16PM +0200, Zeeshan Ali (Khattak) wrote:
> >> On Fri, Nov 30, 2012 at 11:37 AM, Christophe Fergeau
> >> <cfergeau redhat com> wrote:
> >> >This API seems to be very fedora-specific,
> >>
> >> The format is more like Linux specific but yeah, currently we only
> >> need this for Fedora and 'keyboard' command is required in the
> >> kickstart.
> >
> > Ubuntu (and Debian as well I guess) uses a similar but different format,
> > 'layout_variant' as opposed to the 'layout (variant)' that you suggest
> > using in the public API.
> 
> We can easily translate it to their format then?

My point was that this is not a 'linux-specific' format as you said, but a
fedora-specific one.

> 
> > XkbLayout seems to be quite close to a country
> > code, so is easy to get from a 'gettext' format.
> 
> I don't know if i got this. You mean we can guess/derive layout from
> l10-language param?

Yes, l10-language is xx_COUNTRYCODE, and the xkblayouts seem close to that
most of the time (?), but you need to be able to specify a variant as well
(eg some countries have qwertz and qwerty variants).

> > And I'd prefer if we spent a bit of time
> > trying to get the API right now.
> 
> The API is already there, I'm just trying to decide/fix a format for
> the argument/param involved. I'm afraid you have missed the best
> opportunity for discussing these APIs in detail: when they were
> proposed by Daniel for the first time many months ago.

Ah, my bad, I didn't pay attention that we already had made a release with
this function in, I assumed you were adding it. On the other hand, the
format of the argument must have already been decided/fixed when this was
committed, there shouldn't be any decision to make there? Or was it totally
unused so far?

Christophe

Attachment: pgpKRxB7kF9mm.pgp
Description: PGP signature


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]