BES island flags

Doug Ewell doug at
Sat Sep 26 14:59:41 CDT 2020

Christoph Päper wrote:

> Unicode first needed to define a completely new method involving
> otherwise unused Tag characters to encode ISO 3166-2 codes as emoji
> flags. This enabled _thousands_ of new valid sequences, not all of
> which have a definitive flag design associated. Of these, only _three_
> were RGIed then.

I don't consider it as a problem that not every country subdivision has its own flag. If someone tries to encode, say, the flag of one of the provinces (wilayas) of Algeria, which AFAICT do not have their own flags, they will get a fallback representation, which is pretty much what they asked for. Most UIs do not offer a non-existent flag as an option anyway.

> (Few years before, a dozen of preexisting emoji flags lead to the
> systematic encoding of about two hundred geographic codes, some of
> which have no unique flag associated with them.)

Again, I don't see this as a breaking problem. Most of the alternative solutions have their own problems; for example, if a system excludes the flag of Svalbard and Jan Mayen because it is the same as the flag of Norway, then the system (not just the font) has to be revised if a separate flag is introduced later.

I would have considered it very much a problem, and said so at the time, if Unicode had chosen to encode only the ten flags (CN DE ES FR GB IT JP KR RU US) that Japanese cell phone vendors had felt it sufficient to implement.

> The only preexisting implementation (WhatsApp) used a different method (RIS
> ‘XE’ etc.) which was absolutely sufficient for handling just a handful of
> flags.

But doesn't scale beyond just a handful. What if I want to represent the highly distinctive and popular flag of Bavaria (DE-BY)? What about Québec or Tokyo?

> UTC created a kind of Mexican standoff thereafter: they expect vendors
> to show implementation interest in certain flag emojis before formally
> recommending any further ones, while vendors are waiting for the
> consortium to recommend certain codes for implementation before taking
> action.

I do agree with this concern about RGI. The usual response is that the member companies of Unicode who determine what should be recommended are the ones who would be implementing them. But of course, not all major vendors are members of Unicode, and in any case there seems to be reluctance to touch this list regardless of evidence of usage.

>> I question that the only use for subdivision coding in CLDR or in emoji flags was to distinguish England, Scotland, and Wales.
> Such as …?

CLDR provides an extension (U) to identify language or other locale information by, among other criteria, country subdivision. For example, Canadian French (fr-CA) can be sub-identified as either Québec French (fr-CA-u-sd-caqc) or Acadian French (fr-CA-u-sd-canb).

As for emoji flags, as stated above, other country subdivisions besides England, Scotland, and Wales are of local or global interest. People unfamiliar with CLDR extension U complain frequently on social media platforms that they need an emoji flag for their state, province, region, department, or oblast.

Doug Ewell, CC, ALB | Thornton, CO, US |

More information about the Unicode mailing list