Browser IDN display policy: opinions sought
michel at suignard.com
Sat Dec 10 19:20:29 CET 2011
> On 10 dec 2011, at 18:26, Paul Hoffman wrote:
>> D: Unicode if the label is a single script that is displayable by the
>> browser, Punycode otherwise.
>With the exceptions for combinations of various scripts and script COMMON.
In fact, mixed scripts are fine and desired in many situations. Think of Romaji in Japan which cannot be confused with Kana and Kanji. So in that case it is perfectly OK to have a white list of scripts that can be mixed with Latin (which is the typical case). You really want to block the obvious threats such as Latin/Cyrillic and Latin/Greek but let mixed scripts when there is no confusability. It is a case though where you want a white list instead of a black list because new scripts get added to Unicode (and eventually to IDN) which are highly confusable with other scripts.
>Or in other words: If the domain name can be displayed as a U-label,
> in a technically safe way, why not display it as an U-label?
>If I, that do not know Han, get a label in Han, I claim it is as
> non-understandable by me in Han as an A-label, but I must
>say it is much more nice to see it in Han than the alternative.
>Why discriminate against some scripts?
> I.e. my view is that we should really limit the number of cases
> where A-label is displayed to the user.
Agree, I also think that most 'type A' implementations are too restrictive in that aspect.
>If nothing else because people should and must start to
> understand we live in a multi script and multi cultural world.
More information about the Idna-update