On (12/23/2009 12:36 PM), FU Ning-san wrote:
> on 2009?12?23? 10:55, Takao Fujiwara wrote:
>> On (12/23/2009 11:26 AM), FU Ning-san wrote:
>>> Does the man page mention "locale.alias" file and "Latin alphabet"
>>> locale somewhere?
>>> By default the traditional "C/POSIX" locale is not displayed on the
>>> language selection list. It can be enabled by uncomment the line in
>>> /usr/share/gdm/locale.alias file. But the name displayed on the list
>>> becomes "Latin alphabet".
>>
>> It might be my failure.
>> The locale.alias is not upstreamed yet but under discussion.
>> Probably I think it's better to revert "Latin alphabet" to "Unspecified".
>>
>> Currently the name of "C/POSIX" is not good since it's not the
>> language name in GDM language list.
>> My suggestion was "Latin alphabet" to modify the original
>> "Unspecified" to follow the explanation of wikipedia.
>> But it's also not good because C locale doesn't mean either latin only
>> or English only.
>> Since currently there is no good explanation, probably it would be
>> better to revert to the original "Unspecified".
>>
>
> Hi, Takao:
>
> "Unspecified" and "Latin alphabet" are not language names, either. Why
> not just use "C/POSIX"?

My understanding is "Unspecified" means unspecified language.
As I noted above, "C/POSIX" is not a language name but a locale name.

> I think for normal English speaking users they would be fine with
> en_US.UTF-8 and won't bother to look for "C/POSIX". "C/POSIX" only has
> meaning to those who know it -- such as developers. I think for those
> people "C/POSIX" is much straightforward than other expressions.

However some people(not me) doesn't like to show "C/POSIX" since it's 
not a language name and they think showing "C/POSIX" confuse users, 
which language it indicates.
If you compare MS-Windows, it doesn't show any locale names but language 
names only.
So currently it's "C/POSIX" is not a good solution because somebody 
might like but others don't.

Actually I have seen several bugs about this on the upstream bugzilla.
My suggestion "Latin alphabet" was a compromise idea for C vs 
Unspecified but it's still not good.
Currently the recommendation is en_US.UTF-8 for end users and C would be 
a workaround for admins/developers/testers and "Unspecified" could be 
used for the purpose.
So probably we might like to suggest en_US.UTF-8 at first and if it 
doesn't meet the usage, "Unspecified" would be shown as a workaround.

If you'd like the furthermore discussions, the following bug might be good:
https://bugzilla.gnome.org/show_bug.cgi?id=592293

Thanks,
fujiwara

>
> Just my two cents.
>
> Best Regards,
> Harry Fu

Reply via email to