Re: ace leaking (was: Re: [idn] Debunking the ACE myth)

[Martin Duerst <duerst@w3.org>]

At 11:29 01/07/24 -0700, Paul Hoffman / IMC wrote:
>At 12:10 PM -0400 7/24/01, Marc Blanchet wrote:
>>I would guess the opposite (i.e. two domain names: one pure ascii and one 
>>full-idn (without showing the ace equivalent) ,
>>but who knows the future...
>
>That's easy for someone whose company name makes reasonable sense in pure 
>ASCII to say. Telling Chinese and Indian speakers that they must register 
>a domain name in pure ASCII is quite different. Romanization works poorly 
>for many languages; few Americans and Europeans realize that.

Yes, romanization works quite purely for some languages. Many Chinese
sites for example use mostly numeric names. But even where romanization
(or whatever else is currently used in these places) works purely, it will
work much better than openly exposed ACE, in many if not most cases.
Of course that doesn't mean everybody should be forced to get an ASCII
name, but we should at least assume that many will.

Regards,   Martin.