Re: [art] Modern Network Unicode — –02 submitted

Saying U+FFFE and U+FFFF (or any other noncharacters) “MUST NOT be used” merely “as per the Unicode specification” ought to be guided by looking at what the intent of Unicode actually is.

See corrigendum 9 of the Unicode Standard for that intent: http://www.unicode.org/versions/corrigendum9.html

“Noncharacters in the Unicode Standard are intended for internal use and have no standard interpretation when exchanged outside the context of internal use. However, they are not illegal in interchange nor do they cause ill-formed Unicode text. This has always been the intent of the standard, as expressed by the Unicode Technical Committee.”

Since noncharacters have no standard meaning outside of internal use, however, they may be even more problematic than the Unicode paragraph and line separators (which do have standard meaning but are forbidden in the draft for Modern Network Unicode).  Many kinds of protocol strings, such as URIs, IRIs, and strings complying with the PRECIS framework, do not allow noncharacter code points, while other kinds of UTF-8 text, such as JSON, do allow noncharacters.  XML allows all noncharacters except for U+FFFE and U+FFFF.

--Peter

From: Carsten Bormann
Sent: Monday, July 8, 2019 4:24 PM
To: Manger, James
Cc: art@ietf.org
Subject: Re: [art]Modern Network Unicode — –02 submitted

Hi James, Tim, Martin,

thank you for the quick feedback!  As suggested, I have started turning this into a standalone document (of course, still requiring RFC 3629, the UTF-8 definition):

Status:         https://datatracker.ietf.org/doc/draft-bormann-dispatch-modern-network-unicode/
Htmlized:       https://tools.ietf.org/html/draft-bormann-dispatch-modern-network-unicode-02
Diff:           https://tools.ietf.org/rfcdiff?url2=draft-bormann-dispatch-modern-network-unicode-02

I hope I haven’t missed anything important from RFC 5198 that I wanted to keep.  
Maybe time to involve the authors of that RFC…

Grüße, Carsten

> On Jul 8, 2019, at 04:18, Manger, James <James.H.Manger@team.telstra.com> wrote:
> 
> Would be nicer if it wasn't written as a diff from RFC5198 (Network Unicode). That is, if you could get all the rules directly from this doc. For instance, I assume Clean Modern Network Unicode must/should be NFC. Keep the RFC5198 comparisons for an informative annex.

_______________________________________________
art mailing list
art@ietf.org
https://www.ietf.org/mailman/listinfo/art