Le 27/09/2017 à 19:02, Melinda lyons a écrit :
> Dear JAC Members,
>
> ISO 639 RAs-JAC ballot 2017-1 : Montenegrin
>
> Please submit your ballot no later than 10 October 2017.
>
==
Dear All,
Thank you for all comments – I read it attentively. All in all, I am
inclining in favor of Montenegrin having its own code, as long as
Croatian, Serbian and Bosnian cannot be "undone" having their own codes.
Montenegrin has its specific letters ś and ź that cannot be included
under the umbrella "Serbian", or if yes, then perhaps it should be
renamed Serbian-Montenegrin.
I recognize that the former Yugoslavia’s languages started more a
political concept rather than a linguistic one at the time of Yugoslavia
split. The Yugoslavia’s divisions lasted several years, each new country
getting its language code with the exception of the last one,
Montenegro. In addition the last split of two countries is also between
two scripts, Cyrillic and Latin. Today with more and more databases
relying on country codes and language codes the lack of language code of
a sovereign country implies difficulties and may become a security problem.
Elisabeth Porteneuve
==
My vote:
==
Dear JAC Members,
ISO 639 RAs-JAC ballot 2017-1 : Montenegrin
Please submit your ballot no later than 10 October 2017.
A – Inclusion
A1 – Alpha-3
_X_ I am in favor of including Montenegrin in the alpha-3 language code
of ISO 639 Parts 2 and 3.
___ I am opposed to including Montenegrin in the alpha-3 language code.
B – Identifier(s)
_X_ I am in favor of the identifiers cnr (provided that the item will be
encoded).
___ I am opposed to this identifier. [If you have answered “opposed” to
question A2, you don’t need to express opposition to the alpha-2
identifier explicitly.]
C – Names
_X_ I am in favor of the language names Montenegrin (for English),
monténégrin (for French), црногорски jeзик / crnogorski jezik (as
indigenous name).
___ I am opposed to these names. [Please comment!]
==
ISO3166 STANDARD AS A DATABASE
Since the ISO3166 standard as database (we migrated our Word documents
at the end of 2013) we check consistency across all elements, and we do
it with computer tools, not with eyes. Besides codes, any name within
the ISO3166 standard (countries, territories, subdivisions) has language
attributes: language code, and Romanization code, if applies. The
language attributes are necessary for the very precise reason - a word
is meaningful if we know the language in which it is written.
Example of Romanization and native Latin script:
Serbian is in Cyrillic:
"RS","SRB","sr","srp","ROMZ","Serbian Cyrillic (1977)"
Montenegrin is in Latin:
"ME","MNE","","001","LATN",""
Moldova uses Romanian, in Latin:
"MD","MDA","ro","ron","LATN",""
The principal tables in the ISO 3166 database are:
==> subdivision-categories.csv <==
alpha_2_code,alpha_3_code,numeric_code,category_id,language_alpha_2,language_alpha_3,category_name,category_name_plural
==> subdivision-names.csv <==
alpha_2_code,alpha_3_code,numeric_code,subdivision_category_id,subdivision_code,language_alpha_2,language_alpha_3,subdivision_name,romanization_system
==> subdivisions.csv <==
alpha_2_code,alpha_3_code,numeric_code,subdivision_category_id,subdivision_code,subdivision_footnote,subdivision_parent
==> territories.csv <==
alpha_2_code,alpha_3_code,numeric_code,territory_id,language_alpha_2,language_alpha_3,territory_name
For each language in Latin script, or its Romanization if applies, the
ISO3166 database have a set of “allowed characters” (U+abcd, visual,
UTF-8, Unicode universal name). Such set of characters is composed of
[a-z, A-Z] plus diacritics.
Example of “allowed characters” in Romanian:
U+021A Ț c8 9a LATIN CAPITAL LETTER T WITH COMMA BELOW
U+021B ț c8 9b LATIN SMALL LETTER T WITH COMMA BELOW
The check for consistency of the ISO3166 database shall be as automatic
as possible, based on country-codes and then on relevant language-codes.
For each (alpha_2_code,alpha_3_code) verify if all names in
(language_alpha_2,language_alpha_3) satisfy the corresponding characters
set.
That procedure fails if we do not have any of language codes (alpha_2 or
alpha_3), i.e. if both are empty.
Example of how bad it is: the OBP website for Montenegro “hides the
problem” https://www.iso.org/obp/ui#iso:code:3166:ME
cf. line Administrative language(s) with ‘001’ but no language code in
any of Montenegro’s subdivision. Note that any customer subscribing to
the database content (SQL) for his needs, and trying to use automatic
tools, can see there is something broken. Amongst customers are private
and public sector, various national administrations, specialized
agencies, including security ones.
Amongst observers of anomalies are bad boys.
There is no good choice to fix it outside of fundamentals: setting
language code.
==
|