ietf-822
[Top] [All Lists]

Re: Internationalization of the Internet

1994-11-18 23:39:01
On Sat, 19 Nov 1994, Masataka Ohta wrote:

And now is the time to consider the Internaionalized Internet including
mails, to which MIME is unrelated.

Just to jump to the defence of Ohta-san here: I've had fairly lengthy
discussions with him in private e-mail, and it seems that he is positing a
complete overhaul of text on the Internet, whether it be in mail messages,
the DNS, whois servers, or whatever, to put it into an international
context.  Most of those services do not have the ability to mark charset
as MIME does.  Hence ISO-2022-INT-1 which has the marking internal to the
encoding rather than passed around as an out-of-band parameter, and which
can pass transparently through most existing Internet software which
handles text. 

Now, in the context of MIME, (I think) he believes that once this
internationalization goal is acheived, the need to mark charset is
ludicrous, and maybe we should be heading towards a world where the
default is ISO-2022-INT-1 and not US-ASCII.  He himself doesn't use MIME
because he doesn't see the need for it. 

For myself, I think getting rid of the labels we have carefully crafted
into MIME and/or changing the charset default is a bad idea.  The Internet
won't change overnight to use ISO-2022-INT-1, UNICODE, or any other
charset that may be posited in the future for internationalization.  Also,
the application of internationalized charsets in other protocols is
outside of the scope of this mailing list, and removing labels from MIME
merely because other standards don't have labels is not a good design
decision.

Can we stop now please?  It's getting a little tiring.  I appauld the
Swedish efforts to standardise on ISO 8859/1.  It may not be optimal for
internationalization in the long run, but in the short term it is going to
solve a lot of nagging problems in that particular community. 

Cheers,

Rhys.