[uf-discuss] n optimization internationalization (Was: HTML5 support)

Scott Reynen scott at randomchaos.com
Thu Jul 22 21:41:41 PDT 2010

On Jul 22, 2010, at 1:51 PM, Angelo Gladding wrote:

> On Thu, Jul 22, 2010 at 7:20 AM, Oli Studholme
> <microformats.org at boblet.net> wrote:
>> 宮野衆 is the Japanese name Miyano (宮野) Shu (衆) (well, probably — there
>> may be other readings for 衆). As Philip correctly guesses, Miyano is
>> the family name, so inserting any form of space character would give
>> an incorrectly reversed name using implied “n” optimisation.
> My original intentions were to fall back on @lang in case sniffing
> Unicode ranges couldn't
> handle all of the cases. However, if that were the case, would it too
> be sufficiently magic?
> As I mentioned to Philip above, I'll draft the algorithm and post it
> back to be more clear.

I don't believe any algorithm can reliably predict how n optimization should be applied, so it should be used sparingly (only when name order is known) even with increased consideration of non-English names.

I know plenty of Japanese people who, at least when they're interacting primarily with English speakers, write their name given name first (e.g. Shu Miyano), just as most English speakers do.  Sometimes they even do this when writing their names in Japanese.  A couple examples:


Note that both names are printed both ways, given name first and family name first.  Although they can be useful for making better guessing, neither language nor unicode ranges can reliably tell us which name is given and which is family.


More information about the microformats-discuss mailing list