W3C home > Mailing lists > Public > uri@w3.org > April 2002

Re: URIs which identify multiple resources?

From: Mark Nottingham <mnot@mnot.net>
Date: Thu, 4 Apr 2002 13:13:07 -0800
Cc: uri@w3.org
To: Michael Mealling <michael@neonym.net>
Message-Id: <C280DE36-4810-11D6-9446-000A27836A68@mnot.net>

On Thursday, April 4, 2002, at 12:49  PM, Michael Mealling wrote:
> The first paragraph of the Introduction in RFC2368 not withstanding, I
> think both documents identify an action ("send mail this way", or "send
> an sms this way") instead of actual mailboxes or end points.  I would
> prefer documents to be precise about what they actually identify but
> as long as they do so and people use the identifier in a way consistent
> with that statement then I think its ok....

I'm a bit surprised by this; I had though that it was agreed that mailto 
specifically does not specify an action, but only identifies a mailbox; 
it's up to the consumer of the URI to determine what to do with it when 
it's dereferenced.

> In other words, the SMS document has it right: the URI here identifies
> and "SMS message" not the end point for an SMS message. An end point is
> needed as part of an "SMS message" in order to be meaninful but it isn't
> the end point that's being identified.

Where does it say this? The abstract says

    This memo specifies a URI (Universal Resource Identifier) scheme
    "sms" for specifying a recipient (and optionally a gateway) for an
    SMS message.

I'd think that identifying messages is best left to mechanisms like 

> In many cases people have been using mailto:foo@bar.com to identify the
> human user who currently 'uses' that mailbox and IMHO, that's also a 
> serious
> semantic mistake...

I'd agree with that; one should say 'the person who uses the mailbox 
foo@bar.com'; e.g.,

[ foaf:mbox <foo@bar.com> ]

Mark Nottingham
Received on Thursday, 4 April 2002 16:13:11 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:04 UTC