Question regarding LC_MONETARY definitions for fr_CH and de_CH

Question regarding LC_MONETARY definitions for fr_CH and de_CH

[Jakob Runge]

[-- Attachment #1.1: Type: text/plain, Size: 1421 bytes --]

Hello,

given that I'm new to this I hope that I don't tread on anyone's toes
and have found the right place for my question.
If not I'd be happy for any advise.

I've recently stumbled [1] over the LC_MONETARY definitions for fr_CH
and because fr_CH copies these from de_CH also over the ones from de_CH.
If I'm not completely mistaken these seem to be located in
'localedata/locales/fr_CH' and 'localedata/locales/de_CH' respectively.

When comparing the 'p_cs_precedes' and 'n_cs_precedes' settings against
the pages [2,3] from the unicode cldr survey-tool [4] I found what to
appears to be a mismatch to me for fr_CH.
That is for fr_CH the currency format in the CLDR appears to be defined
as '#,##0.00 ¤' which would be '1 295.00 CHF' or '-1 295.00 CHF',
but for 'localedata/locales/fr_CH' the values of 'p_cs_precedes' and
'n_cs_precedes' would be 1 on the current master commit (1f50f2a).

If my observations are correct it would seem to me that the
'p_cs_precedes' and 'n_cs_precedes' for fr_CH should be adjusted.

I'd love to read confirmation or correction on this.

All the best,
  Jakob


[1]: https://stackoverflow.com/q/56057751/448591
[2]:
https://st.unicode.org/cldr-apps/v#/fr_CH/Number_Formatting_Patterns/24a93b3d14ba17b2
[3]:
https://st.unicode.org/cldr-apps/v#/de_CH/Number_Formatting_Patterns/53687a25c19b6481
[4]: http://cldr.unicode.org/index/survey-tool


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: Question regarding LC_MONETARY definitions for fr_CH and de_CH

[Carlos O'Donell]

On 5/17/19 7:12 PM, Jakob Runge wrote:
> If my observations are correct it would seem to me that the
> 'p_cs_precedes' and 'n_cs_precedes' for fr_CH should be adjusted.
> 
> I'd love to read confirmation or correction on this.

The current data in fr_CH was contributed by Keld Simonsen, and you
should reach out to him also to discuss this issue.

In general glibc tries to match published government standards about
language locales and particularly for number formats.

Quoting that CLDR is different is not a good justification. Yes, we
are trying hard to harmonize the glibc locales with CLDR since it
avoids confusion. However, we also need to understand which database
has it right, and correct the other.

Are you a Swiss resident? Do you have experience with the locale
in question?

-- 
Cheers,
Carlos.

Re: Question regarding LC_MONETARY definitions for fr_CH and de_CH

[Jakob Runge]

[-- Attachment #1.1: Type: text/plain, Size: 1257 bytes --]

> The current data in fr_CH was contributed by Keld Simonsen, and you
> should reach out to him also to discuss this issue.

Thanks for the pointer - I will do that!

> In general glibc tries to match published government standards about
> language locales and particularly for number formats.
> 
> Quoting that CLDR is different is not a good justification. Yes, we
> are trying hard to harmonize the glibc locales with CLDR since it
> avoids confusion. However, we also need to understand which database
> has it right, and correct the other.

I understand that this makes perfectly sense and would also like to
understand the source for this both in CLDR as well as glibc.

> Are you a Swiss resident? Do you have experience with the locale
> in question?
> 

No - I'm not a Swiss resident, and can't speak french - so I'm fairly
removed from fr_CH. I tried to look this up on sites of the Swiss
national bank and similar institutions but had no luck in my research.
To the extend that I could get information from Swiss people it seemed
to be conflicting and unclear whether it related to fr_CH in particular.

So no - I wouldn't say that I have particular experience with the locale
in question.

All the best,
  Jakob


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: Question regarding LC_MONETARY definitions for fr_CH and de_CH

[Jakob Runge]

[-- Attachment #1.1: Type: text/plain, Size: 2448 bytes --]

On 5/18/19 6:22 AM, Carlos O'Donell wrote:
> On 5/17/19 7:12 PM, Jakob Runge wrote:
>> If my observations are correct it would seem to me that the
>> 'p_cs_precedes' and 'n_cs_precedes' for fr_CH should be adjusted.
>>
>> I'd love to read confirmation or correction on this.
> 
> The current data in fr_CH was contributed by Keld Simonsen, and you
> should reach out to him also to discuss this issue.
> 
> In general glibc tries to match published government standards about
> language locales and particularly for number formats.
> 
> Quoting that CLDR is different is not a good justification. Yes, we
> are trying hard to harmonize the glibc locales with CLDR since it
> avoids confusion. However, we also need to understand which database
> has it right, and correct the other.

I've dug deeper and have expended my answer at StackOverflow [1].
The CLDR ticket [2] that discusses this formatting change cites, among
other sources, the CERN guide on typography [3].

This guide specifies:

> Écriture des sommes d’argent
> 
> Le nombre s’écrit par tranches de trois chiffres séparées par un espace insécable
> (pas de point ni d’apostrophe de séparation), et est suivi (et jamais précédé) de
> l’indication de la monnaie soit au long, soit en abrégé. Pour les abréviations de noms
> de monnaie, on utilise le code ISO.

Using google translate this roughly translates to:

> Writing sums of money
> 
> The number is written in three-digit increments separated by a non-breaking space
> (no point or apostrophe of separation), and is followed (and never preceded) by
> the indication of the currency is long or abbreviated. For name abbreviations
> currency, we use the ISO code.

From this I gather that indeed the `p_cs_precedes` and `n_cs_precedes`
settings for fr_CH should be adjusted.

Since I don't have further information on other *_CH locales, and the
settings for de_CH seem to coincide with the current CLDR settings it is
my believe that only fr_CH is affected by this and should therefore be
different from de_CH and not have:

> LC_MONETARY
> copy  "de_CH"
> END LC_MONETAR

All the best,
  Jakob

[1]: https://stackoverflow.com/a/56161892/448591
[2]: https://unicode-org.atlassian.net/browse/CLDR-9370
[3]:
https://translation-council-support-group.web.cern.ch/sites/translation-council-support-group.web.cern.ch/files/Guide%20Typo_FR_styleCERN.pdf


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]