Re: An smtp URL scheme

Walter Ian Kaye (
Thu, 10 Jul 1997 09:27:07 -0700

Message-Id: <v03102829afeabcb8f20a@[]>
In-Reply-To: <>
Date: Thu, 10 Jul 1997 09:27:07 -0700
From: Walter Ian Kaye <>
Subject: Re: An smtp URL scheme

At 9:41a -0400 07/10/97, Greg Marr wrote:
 > At 10:04 PM 7/9/97 -0700, Walter Ian Kaye wrote:
 > >At 9:51a -0400 07/09/97, Greg Marr wrote:
 > > > However, the preferred way is
 > > > smtp://host/user?subject=subject+body=yes+X-header=header
 > >
 > >This makes no sense. CGI's expect ampersands as delimiters.
 > Except for the ones that expect +'s. with an encoding type of x-www-
 > encoded or something like that instead of x-url-encoded.  There was
 > an entire thread or two about this on one of the mailing lists I am
 > on a while ago.  I'll have to track it down sometime.  I didn't
 > invent this, I highly double that Richard Levitte and I both invented
 > the same syntax out of thin air.

If you do track it down, I'd be interested in seeing it. The documentation
at speaks only of ampersands...

 > > > because otherwise the URL should be encoded to:
 > > > smtp://host/user?subject=subject&amp;body=body&amp;X-
 > header=header
 > > > when it is included in an HTML page.
 > >
 > >No, you are confusing rendered text with attribute values.
 > How am I confusing rendered text with attribute values?
 > Are you saying that entity replacement shouldn't be performed on
 > attribute values?

I'm saying it is not necessary to use entities for ampersands within
an attribute value.

  Walter Ian Kaye <boo_at_best*com>    Programmer - Excel, AppleScript,
          Mountain View, CA                         ProTERM, FoxPro, HTML     Musician - Guitarist, Songwriter