RE: Proposal to register SCSU for IANA

Hi Markus and Martin,

I just visited the Unicode Consortium home page and verified
my memory - UTF#6 (SCSU) has the status 'Approved' (by
Unicode Consortium) but is NOT incorporated into the
brand new Unicode v3.0 specification.  I *think* that
weakens the case for IANA registration, but I could easily
be wrong.

Harald or Ned - what do you guys think?

Cheers,
- Ira McDonald, consulting architect at Sharp Labs America
  High North Inc

-----Original Message-----
From: Martin J. Duerst [mailto:duerst@w3.org]
Sent: Monday, March 20, 2000 3:23 PM
To: Markus Scherer; charsets@innosoft.com
Subject: Re: Proposal to register SCSU for IANA


This is probably a difficult one.

- We don't really want it to be used too much, but the IANA
   policy for registrations (as far as I have heard from people
   I consider relevant in this respect) currently tends towards
   'register if somebody uses it, even if we think it shouldn't
   be used'.

- Strictly formally speaking, the thing fits the definition for
   'charset', but it's really rather far away from the average
   case.

So I would say: Register, but with as big a warning as possible.


Regards,   Martin.

At 00/03/20 10:49 -0800, Markus Scherer wrote:
>Hello,
>
>I just subscribed to this list a few days ago and did not see any traffic 
>that would have shown me the table manners.
>My job is with IBM in a team that produces Unicode- and I18N-related 
>libraries for Java and C/C++.
>
>Proposal: I would like to register an encoding as follows with the IANA 
>list at
>http://www.isi.edu/in-notes/iana/assignments/character-sets :
>
>
>Name: SCSU
>MIBenum: 107 ??
>Source: A Standard Compression Scheme for Unicode, Unicode Technical 
>Report #6, http://www.unicode.org/unicode/reports/tr6/
>Alias: [none]
>
>
>This is a character encoding scheme for Unicode. Algorithmic 
>transformation of Unicode text to and from SCSU is fairly
>simple and can yield a similar code point:byte count ratio as legacy 
>encodings. It is stateful. US-ASCII and ISO-8859-1 text
>that does not contain control codes other than NUL, TAB, CR, LF is also 
>correct SCSU text. SCSU uses all byte values.
>
>
>I am looking forward to get feedback from you and hope to advance this 
>towards a registration.
>
>
>Thanks,
>
>markus

Received on Monday, 20 March 2000 22:50:38 UTC