RE: Small comment to UTF-7 draft

The best solution is to support only one form of UTF- in each 
application domain.
Clearly, UTF-7 is meant for mail transport. So it should become the 'only' UTF-
supported in Internet Mail context. UTF-8 is file system safe, and has 
other nice
properties that make it attractive to solve a particular set of 
problems in the UNIX
system environment. It should be the 'only' such UTF- supported in that domain.

We should not even consider applying two UTF-s to the same domain.
Finally, we should remember what the raison d'etre is for these forms: 
a temporary
bridge that works with current technologies. Flat 16-bits remains the long term
preferred target encoding for ALL application domains.
A./
----------
| From: David Goldsmith  <netmail!David_Goldsmith@taligent.com>
| To: IETF Charsets  <ietf-charsets@INNOSOFT.COM>;  <unicored@Unicode.ORG>
| Subject: Small comment to UTF-7 draft
| Date: Monday, May 16, 1994 3:05PM
|
| Received: from UNICODE.ORG by netmail.microsoft.com with SMTP (5.65/25-eef)
| 	id AA26548; Mon, 16 May 94 15:23:19 -0700
| Received: from taligent.com by Unicode.ORG (NX5.67c/NX3.0M)
| 	id AA19788; Mon, 16 May 94 14:46:00 -0700
| Received: from david-goldsmith.taligent.com by taligent.com
| with SMTP (5.67/23-Oct-1991-eef)
| 	id AA14169; Mon, 16 May 94 15:05:50 -0700
| 	for
| Message-Id: <9405162205.AA14169@taligent.com>
| X-Sender: dgold@banpeikun-rx.taligent.com
| Mime-Version: 1.0
| Content-Type: text/plain; charset="us-ascii"
|
| Below is a private mail message I received recently. I've received the
| author's permission to redirect it to these two mailing lists for
| discussion. I do not currently have a position on this issue and I wanted
| to hear from the community.
|
| >Date: Tue, 10 May 94 11:19:37 +0200
| >From: Dan Oscarsson <Dan.Oscarsson@malmo.trab.se>
| >To: david_goldsmith@taligent.com
| >Subject: Small comment to UTF-7 draft
| >X-Sun-Charset: US-ASCII
| >
| >Hi
| >
| >I just read the utf7-03 draft for UTF-7 and saw one thing I would like to
| >comment on.
| >
| >In the summary you recommend UTF-7 only to be used with 7 bit 
transports, but
| >I would very much recommend to use UTF-7 with 8 bit transports if the text
| >is a mainly latin character based text. Then all ISO 8859-1 characters can
| >be sent as themselves (as ISO 8859-1 is a true subset of UCS-2) and only
| >those character outside ISO 8859-1 to be encoded. This is much better than
| >using
| >UTF-8 which is NOT ISO 8859-1 compatible and will for all users that use
| >ISO 8859-1 as their standard (and ISO 8859-1 is the defacto standard today)
| >look like garbage. UTF-7 on a 8 bit transport with ISO 8859-1 as character
| >coding will works just like quoted-printable, but will make most characters
| >in ISO 8859-1 readable even for users without a viewer that understands
| >the UTF-7 encoding.
| >
| >I suggest strongly that UTF-7 should recommend it to be used with 8 bit
| >transports for all texts with mainly latin script.
| >
| >Regards,
| >
| >    Dan
| >
| >--
| >Dan Oscarsson
| >Telia Research AB                       Email: Dan.Oscarsson@malmo.trab.se
| >Box 85
| >201 20  Malmo, Sweden
| >
|
| ----------------------------
| David Goldsmith
| david_goldsmith@taligent.com
| Taligent, Inc.
| 10201 N. DeAnza Blvd.
| Cupertino, CA  95014-2233
|
|
| 

--Boundary (ID uEbHHWxWEwCKT9wM3evJ5w)

Received on Monday, 16 May 1994 16:27:50 UTC