W3C home > Mailing lists > Public > www-html@w3.org > April 2003

Re: 'email' element type

From: Etan Wexler <ewexler@stickdog.com>
Date: Tue, 08 Apr 2003 17:04:44 -0700
To: "Todd O'Bryan" <toddobryan@mac.com>, www-html@w3.org
Message-id: <BAB8A197.D3E%ewexler@stickdog.com>

Todd O'Bryan wrote to <mailto:www-html@w3.org> on 8 April 2003 in "Re:
'email' element type" (<mid:76DC2D17-69BB-11D7-99A8-0030656EC0E4@mac.com>):

> But, isn't "mailto:" kind of a fake URI? I mean it doesn't really
> describe a protocol and a link so much as tell a browser to use an
> email program and stick something in the To: box.

Though I consider "mailto" an unfortunate misnomer ("mailbox" would have
been better), the "mailto" scheme is in no way fake. URI schemes do not need
to correspond to protocols; the "urn" scheme has no defined protocol. But if
we insist on correspondence to protocols as the measure of legitimacy,
"mailto" correponds to Simple Mail Transfer Protocol [SMTP] and the Domain
Name System [DNS].

I'm not sure what you mean by "link". I think that you mean a document, but
I'd like a clarification.

In my book, a link is an explicit specification of a relationship.

> The other obnoxious thing about using href for mailto is that it's not
> really a semantic link. Most (all?) other links lead to information.

If you want the "mailto" scheme to lead to information, have your user agent
query the Domain Name System. For example, you come across
<mailto:user@example.net> and it leads to information from
<user.example.net>.
 
> mailto is a roundabout way of asking for information, namely the email
> you want to send to the address you just clicked on.

That is one treatment, and perfectly valid.

> And don't get me 
> started on people who don't write email addresses in the text of their
> mailto: links so that I click on what I think is hypertext and have my
> email program open up when I really wouldn't have clicked it if I'd
> known what I was in for.

If your user agent does not distinguish links using the "mailto" scheme and
you cannot configure your user agent to distinguish those links, that is a
user agent shortcoming.

I work with a similar shortcoming. My WannaBe browser does not handle FTP.
Any "ftp" URIs are passed to my default FTP application. Should we include
an FTP-designating element type so that WannaBe can easily distinguish this
case?


[SMTP]
John C. Klensin, editor.
"Simple Mail Transfer Protocol".
Internet Engineering Task Force standards-track Request For Comments: RFC
2821.
April 2001.
<urn:ietf:rfc:2821>.

[DNS]
Paul Mockapetris.
"Domain Names - Concepts and Facilities".
Internet Engineering Task Force standards-track Request For Comments: RFC
1034.
November 1987.
<urn:ietf:rfc:1034>.

Paul Mockapetris.
"Domain Names - Implementation and Specification".
Internet Engineering Task Force standards-track Request For Comments: RFC
1035.
November 1987.
<urn:ietf:rfc:1035>.

Collectively <urn:ietf:std:13>.


-- 
Etan Wexler <mailto:ewexler@stickdog.com>
Received on Tuesday, 8 April 2003 20:06:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:55 GMT