Re: FW: [WhatsApp Support] Your Request: Windows Phone Client 2.10.523(ticket #7044796)

From: Philippe Verdy <>
Date: Mon, 5 Aug 2013 19:06:00 +0200

Anyway the only sequences that are mapped to regional indicators are for
private extensions of SJIS in Japan, respetively from KDDI and SoftBank

1F1E8 1F1F3;;F3D2;FBB3 # [CN] People's Republic of China
1F1E9 1F1EA;;F3CF;FBAE # [DE] Germany
1F1EA 1F1F8;;F348;FBB1 # [ES] Spain
1F1EB 1F1F7;;F3CE;FBAD # [FR] France
1F1EC 1F1E7;;F3D1;FBB0 # [GB] United Kingdom
1F1EE 1F1F9;;F3D0;FBAF # [IT] Italian
1F1EF 1F1F5;;F6A5;FBAB # [JP] Japan
1F1F0 1F1F7;;F3D3;FBB4 # [KR] Korean
1F1F7 1F1FA;;F349;FBB2 # [RU] Federation of Russia
1F1FA 1F1F8;;F790;FBAC # [US] United States of America

This limited set of "flags" highly suggests that in fact these flags
are used not really to convey a regional information, but only a set
of well known languages used in these countries and spoken/written
internationally i.e. these are frequently visual indicators of the
language, for use in web menus for language selection (even if it is a
common but bad practice):

- Why the British flag or the US flag should be used to designate the
English language?

- And why not the Indian or South African flags ? and why nothing os
appropraite here for the Portuguese language, and which country flag
to use between Portugal, Brasil, Cape Verde, Guinea Bissau, or even
for the minority Portuguese speaking comumunity in Macau ???).

These mappings for roundtrip compatibility with private extensions of
SJIS, which are not even compatible with each other, and not supported
in the DoCoMo extension of SJIS, means that these characters are
already deprecated from the start. And not all letters are used. It is
also very unlikely that they will be used in sequences longer than 1
"country flag", due to the many missing countries (even of the UCS
encoding could allow mapping other country flags, but with stability
problems whose origin in in ISO 3166-1 (but the above country codes
are stable since long in all ISO 3166-1 versions... except UK for
which the *informative* UCS roundtrip mapping chose to use the GB code
and not the legacy UK code).

For many other countries or regions we would need extensions including
longer codes, or version suffixes. The UCS encoding the way it is made
also allows mapping pairs of codes that are npt associated with any
country and that will never be mapped such as [QQ] or [ZZ]. This type
of roundtrip mapping will then nver be used by these private SJIS

Note finaly that if the country codes above are stable, this is not
true of their flags, and some countries even have several flags for
different use (civil, military with navy variants, enseign,
honorific), plus decorations. Their colors and proportion may also
change over time or depending on presentation (e.g. the tricolor flag
of France has been modified so that its first vertical band (blue)
near the hoist is narrower than the the third vertical band (red), so
that the flag seems to have bands of equal width when th eflag is
waving; and the exact color matching was also changed.
I don't remember the details exactly but this is true as well in
Russia, as well as in Japan since WW2; Spain uses several variants of
its flag. If these flags were hanging vertically, they could also
exist with very different proportions or could keep the orientation of
the bands, and the form of the flag may also no longer be rectangular,
with triangular shapes on the bottom floatting side.
For this reason I still approve the fact that Unicode does not
standardize the colors and shapes. But I still think that it should
haev modeled a scheme allowing more precision if needed, as well as
allowing the representation of all countries, including former ones
(and avoiding ambiguities like [CS] between the former Czechoslovakia
and the former Federation of Serbia and Montenegro : for these
distinctions between codes of former country, ISO 3166 defines
4-letter codes (and if needed, it will use digits); but it will be
impossible to track the country to which the former 2-letter code was
mapped if it's not specified with a stable extension in the code such
as [CS:1945] instead of just [CS].
If these extensions were used, we could represent the IOC white flag
or flags of national OC member teams (using their 3-IOC letter code
after a prefix, such as [-IOC:ENG] for the Rugby team of England when
it does not compete in Olymic games).
I am not convinced that defining these extensions would break the
existing private implemetnations in SJIS.
2013/8/5 Markus Scherer <>
> Dear Pradeep,
> The information you got from WhatsApp is wrong. The Unicode Consortium
> does not "design and create Emoji", and support in WhatsApp for the Indian
> flag is entirely up to WhatsApp. Please read the last two questions at
> and work with WhatsApp
> on support for the Indian flag.
> Best regards,
> markus
Received on Mon Aug 05 2013 - 12:11:27 CDT

This archive was generated by hypermail 2.2.0 : Mon Aug 05 2013 - 12:11:28 CDT