Re: http charset labelling

Gavin Nicol (gtn@ebt.com)
Sun, 18 Feb 1996 20:01:09 -0500


From: Masataka Ohta <mohta@necom830.cc.titech.ac.jp>
Message-Id: <199602160930.SAA17385@necom830.cc.titech.ac.jp>
Subject: Re: http charset labelling
To: keld@dkuug.dk (Keld J|rn Simonsen)
Date: Fri, 16 Feb 96 18:30:24 JST
Cc: dupuy@cs.columbia.edu, gtn@ebt.com, uri@bunyip.com
In-Reply-To: <199602141548.QAA24621@dkuug.dk>; from "Keld J|rn Simonsen" at Feb 14, 96 4:48 pm

Keld;

I wrote:

> > So, code mapping proposal which is proven not to work on paper is,
> > IMHO, almost useless. Maybe, URLs in e-mail are rare exceptions.

Again, code mapping proposal which is proven not to work on paper is
almost useless.

> To me it means that the URL is hidden and can be anything so it
> does not matter then if it is encoded in a special charset or not.

On browsers, yes, it is hidden. So, it is meaningless to use
a special charset.

						Masataka Ohta
Date: Sun, 18 Feb 1996 20:01:09 -0500
From: Gavin Nicol <gtn@ebt.com>
Message-Id: <199602190101.UAA09669@ebt-inc.ebt.com>
To: mohta@necom830.cc.titech.ac.jp
Cc: keld@dkuug.dk, dupuy@cs.columbia.edu, uri@bunyip.com
In-Reply-To: <199602160930.SAA17385@necom830.cc.titech.ac.jp> (message from Masataka Ohta on Fri, 16 Feb 96 18:30:24 JST)
Subject: Re: http charset labelling

>> To me it means that the URL is hidden and can be anything so it
>> does not matter then if it is encoded in a special charset or not.
> 
>On browsers, yes, it is hidden. So, it is meaningless to use
>a special charset.

The point is that if I receive a URL, by whatever means, and then I
type it into the "go to" field of any browser, then when it is sent to
the server, the server has no idea what encoding is being used.

The actual indication of the encoding should be hidden from the user,
but it is still important for it to be there, because even for ASCII
names, maybe the user entered it in zenkaku.