RE: [codelist] Fwd: Re: Impact of changing a specification from URI to IRI and XRI

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

At 2006-12-01 09:47 +0000, Paul Spencer wrote:
>The only danger I know of in using non-ASCII characters is if any harm can
>be caused by someone maliciously (or otherwise) replacing an ASCII character
>with a similar-looking non-ASCII character.

But I believe IRI strings still can only have valid Unicode 
characters.  It isn't just an arbitrary binary string.

I just looked up some details in http://tools.ietf.org/rfc/rfc3987.txt

    The above mapping from IRIs to URIs produces URIs fully conforming to
    [RFC3986].  The mapping is also an identity transformation for URIs
    and is idempotent;  applying the mapping a second time will not
    change anything.  Every URI is by definition an IRI.

>This is the basis of many
>phishing attacks. I don't really see how that affects us, but then I also
>don't see how allowing non-ASCII characters helps us significantly either.
>Do we really need our identifiers to be that multi-lingual?

Yes, I absolutely believe that the identifiers do.  In the ISO 
paperwork I see go by my desk the Asian comments about IRI strings 
are raised for a number of specifications in which URI strings are used.

. . . . . . . . . . . Ken

--
World-wide corporate, govt. & user group XML, XSL and UBL training
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/o/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Cancer Awareness Aug'05  http://www.CraneSoftwrights.com/o/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal