Here is why I don't like n3680.
ISO-3166-1 defines two-letter symbols for many "pieces of earth" which
are not independent countries.
For example, there's an ISO-3166-1 symbol for Réunion, an overseas
department of France.
Why would Unicode define a flag for Réunion (RE, also known as FR-RE in
ISO-3166-2) but not for Puy-de-Dôme (FR-63, a metropolitan department)
or Bretagne (Brittany, FR-E, a metropolitan region which has its own
flag:
http://upload.wikimedia.org/wikipedia/commons/7/7d/Flag_of_Brittany.svg)?
If those flags are used for languages, Brittany has its own language:
Breton (Brezhoneg). So, a support for ISO-3166-2 could be a good Idea.
Also, the flag of Québec (CA-QC,
http://upload.wikimedia.org/wikipedia/commons/5/5f/Flag_of_Quebec.svg)
could be used to make a difference between European and Canadian French.
Unicode is supposed to be a stable standard: any new character will
remain always and forever.
ISO-3166 is not stable enough to be encoded as it is in Unicode:
countries may change their names, countries may merge together,
countried may be divided into several parts.
In any case, new two-letter codes may be created, existing ones may
become obsolete and, most problematic, when a two-letter code has been
obsolete for at least 5 years, IT MAY BE USED FOR ANOTHER COUNTRY.
For these reasons, I think the system proposed by Philippe Verdy would
be better suited to encode flags: the codes would have no meaning by
themselves, but their combinations would be displayed as flags, whenever
possible, by the rendering engine.
Le 01/06/12 18:24, Philippe Verdy a écrit :
This last-chance fallback conversion may be specified with a NFKC
decomposition mapping. For example this<font> compatibility mapping :
XXX00 ; FLAG SYMBOL INITIAL HYPHEN ; ... ; So ; ... ;<font>005B 002D ;
XXX01 ; FLAG SYMBOL INITIAL A ; ... ; So ; ... ;<font>005B 0041 ;
XXX1A ; FLAG SYMBOL INITIAL Z ; ... ; So ; ... ;<font>005B 005A ;
XXX20 ; FLAG SYMBOL INITIAL ZERO ; ... ; So ; ... ;<font>005B 0030 ;
XXX29 ; FLAG SYMBOL INITIAL NINE ; ... ; So ; ... ;<font>005B 0039 ;
...
XXX30 ; FLAG SYMBOL MEDIAL HYPHEN ; ... ; So ; ... ;<font>002D ;
XXX31 ; FLAG SYMBOL MEDIAL A ; ... ; So ; ... ;<font>0041 ;
XXX4A ; FLAG SYMBOL MEDIAL Z ; ... ; So ; ... ;<font>005A ;
XXX50 ; FLAG SYMBOL MEDIAL ZERO ; ... ; So ; ... ;<font>0030 ;
XXX59 ; FLAG SYMBOL MEDIAL NINE ; ... ; So ; ... ;<font>0039 ;
...
XXX60 ; FLAG SYMBOL FINAL HYPHEN ; ... ; So ; ... ;<font>002D ;
XXX61 ; FLAG SYMBOL FINAL A ; ... ; So ; ... ;<font>0041 005D ;
XXX7A ; FLAG SYMBOL FINAL Z ; ... ; So ; ... ;<font>005A 005D ;
XXX80 ; FLAG SYMBOL FINAL ZERO ; ... ; So ; ... ;<font>0030 005D ;
XXX89 ; FLAG SYMBOL FINAL NINE ; ... ; So ; ... ;<font>0039 005D ;
In the following lines, I'll use abbreviations.
For example,
FSIA = FLAG SYMBOL INITIAL A,
FSMHY = FLAG SYMBOL MEDIAL HYPHEN,
FSF9 = FLAG SYMBOL FINAL 9.
When used alone,
FSIA would look like
??
?A
??
FSMA would look like
?
A
?
FSFA would look like
??
A?
??
And FSIA FSMA FSFA would look like
?????
?AAA?
?????
if there's no rendering engine able to convert it into a real flag.
So, the flag of the United States
(http://upload.wikimedia.org/wikipedia/commons/e/e2/Flag_of_the_United_States_%28Pantone%29.svg)
would be "spelled"
FSIU FSFS
It would look like
????
?US?
????
The flag of Nunavut
(http://upload.wikimedia.org/wikipedia/commons/9/90/Flag_of_Nunavut.svg)
would be "spelled"
FSIC FSMA FSMHY FSMN FSFU
It would look like
???????
?CA-NU?
???????
The flag of Chiriquí
(http://upload.wikimedia.org/wikipedia/commons/3/3b/Bandera_de_Chiriqui_%28Panam%C3%A0%29.svg)
would be "spelled"
FSIP FSMA FSMHY FSF4
It would look like
??????
?PA-4?
??????
Etc.
I first thought we could get rid of the hyphen because all the country
symbols are two letters long, but there are four letters long country
codes in ISO-3166-3, therefore removing the hyphen could lead to some
conflicts.
JF
