Date: Tue, 5 Jul 1994 08:22:08 -0400
Message-Id: <199407051222.IAA76382@inca.gate.net>
To: uri@bunyip.com
From: hoymand@gate.net (Dirk Herr-Hoyman)
Subject: Mail URL (was Re: Another snapshot of the URL document)
At 10:04 AM 7/4/94 +0200, Guido.van.Rossum@cwi.nl wrote:
>(on the mailto: URL format)
>
>> So, is this an interface to SMTP
>> mail or is it something else that doesn't have anything to do with SMTP
>> necessarily. Since you have carefully noted ports and RFCs that pertain in
>> other URLs, this seems out of place to me.
>
>The way I understand this, and would like to see it used, and the way
>it is supported by at least on WWW client (Lynx), is that selecting a
>reference to such a URL drops you into your favorite mail program and
>allws you to send a message to the given address. This is useful,
>otherwise a cut-n-paste job would be needed to transfer the email
>address into the mailer.
>
>It would be totally wrong to redefile 'mailto:' to mean an SMTP
>interface -- it represents an email address, not a delivery mechanism.
>
>In a sense, 'mailto:' is similar to 'telnet:' -- once you've selected
>such a URL, you essentially leave the WWW context and start
>interactive at a different level. Unlike most other URL types, these
>URL's don't refer to files that can be mechanically retrieved. But
>since there is a genuine need to include email addresses and telnet
>session parameters in WWW documents, they deserve some level of
>standardization.
>
I understand that mailto represents only a recipient address. And part of
my original message was to get some clarity on exactly what we are trying
to use it for. But, a second point was that there are some resources
available ONLY by e-mail servers, and for these mailto is typically
inadequate. Only mail responders, those addresses typically look like
info@domainname, would work. Most e-mail servers, such as listserv, work
by using a command either in the Subject or body of the message.
This appears to me to be a hole in the URL specification. Even though this
type of a URL would not act synchronously (i.e., you wouldn't get a
document back directly in a Web client from a tcp connection), it
nonetheless would be a means of specifying how to retrieve a resource. As
this draft of the URL spec is more clearly taking the "high ground" and not
tieing itself to any particular client scheme (such as WWW, in particular),
I feel it is an appropriate time to consider the inclusion of such a URL.
Call it mail, if you like, so as to distinguish it from mailto.
-- Dirk Herr-Hoyman <hoymand@gate.net> | Practice CyberBeach Publishing | random acts of kindness * Internet publishing services | and senseless beauty Lake Worth, Florida, USA | Home Page: <URL:http://www.gate.net/~hoymand/cyberbeach.html> Phone: +1.407.540.8309