Re: [hatari-devel] Setting TOS language / keyboard

[Thomas Huth <th.huth@xxxxxxxxx>]

Am Fri, 17 Jun 2022 21:16:09 +0300
schrieb Eero Tamminen <oak@xxxxxxxxxxxxxx>:

> Hi,
> 
> On 12.6.2022 16.43, Thomas Huth wrote:
> > schrieb Eero Tamminen <oak@xxxxxxxxxxxxxx>:  
> >> On 4.6.2022 17.37, Eero Tamminen wrote:  
> >>> Attached are patches to get NVRAM language default from LANG, and new
> >>> "--country" option to set country code for NVRAM-less machines +
> >>> appropriate documentation update.  
> >>
> >> Thomas, Nicolas, any comments, does it look clear enough based on the
> >> documentation?  
> > 
> > Looks reasonable to me at a quick glance. The parsing of $LANG is likely a
> > little bit inprecise (looking at the output of "localedef --list-archive",  
> 
> --------------------------
> $ localedef --list-archive
> en_GB.utf8
> en_US.utf8
> fi_FI.utf8
> --------------------------
> ?

On Fedora, I get:

$ localedef --list-archive | wc -l
866
$ localedef --list-archive | grep "^..._" | head -n 20
agr_PE
agr_PE.utf8
anp_IN
anp_IN.utf8
ast_ES
ast_ES.iso885915
ast_ES.utf8
ayc_PE
ayc_PE.utf8
bem_ZM
bem_ZM.utf8
ber_DZ
ber_DZ.utf8
ber_MA
ber_MA.utf8
bhb_IN.utf8
bho_IN
bho_IN.utf8
bho_NP
bho_NP.utf8

> > there are also locales that use three letters instead of two, for example),
> > but I guess it will be enough for the common languages that are supported
> > by EmuTOS...?  
> 
> I've pushed the LANG commit.
> 
> Matching names for languages supported by EmuTOS is enough.  Not 
> matching languages it does not know about, is not a problem.  I assume 
> there are no XX_LLL language names where first two LL would match EmuTOS 
> supported language code.

I don't get any output for:

 localedef --list-archive | grep "_...\."

.... so I guess that should be ok.

> Any comments on the "--country" option?

Fine for me.

 Thomas