Re: Mail addresses as URLs

Tim Berners-Lee (timbl@www3.cern.ch)
Tue, 11 May 93 17:04:07 +0100


> Date: Tue, 11 May 1993 10:13:49 -0400 (EDT)
> From: John C Klensin <KLENSIN@infoods.mit.edu>
>

> Tim,
>

> I think the idea of permitting an email address is reasonable,
although
> it brings us another step closer to a uniform human locator, as in
"if
> 'mailto' then why not 'faxto', 'telephoneto', 'postto' and so on.

Yes. I guess we have to be a little more definite about the object
which is being located. I prefer to see the web as something with
a state which one can alter, rather than a recipe book scripts behind
the buttons. The
"mailto:" URL suggests the well defined concept of a mailbox which
may or may not contain documents and may or may not have
restricted access. Perhaps "mailbox:" would be better than "mailto:".
Moving a document into somone's mail box is just like moving a
file into a directory: it is an operation with pre- and
postconditions. It just happens that mail uses a particular
protocol which is optimised for particular operations,
has a wide penetration, etc., but can otherwise be treated
with the same user interface metaphor as news and retrieval.

"telephoneto" like "telnet" doesn't fit into that model, and
so is less usefully integrated.

> However you are going to need to be extremely careful about syntax.
The

> correct model should be able to accomodate _any_ valid RFC822
address,
> implying not only "mailto:usermailbox@domain" but including the

> Personal Name <usermailbox@domain>
> form and several others.

Is there not a subsyntax for the usermailbox@domain bit, with the
comment personal Name removed? It would have the example of being
a little closer to something which one can test for equality.

> RFC1327 forms of X.400 addresses projected into the RFC822
environment
> also mean that we need to be very careful about quoting and
structure,
> since those addresses are heavily solidus (/)-laden.

Yes. I had intended to keep "=" in the URL spec
so that it could be used to map /attribute=value/ such as x400.
If you directly quote an RFC822 address, then you run across
the fact MIME uses = instead of %, and all kinds of things.
Would it be possible to remove all RFC822 quoting and apply
URL escaping as a reversable and well-defined transformation
which would presvent the horrible results of layered escaping?

> This makes me quite hesitant about forms like Thomas Bruce's
> mailto://some.mail.host.dom/somebody
> because it implies that we will need to very carefully work out and
> document transformation rules for all of the different variations
of
> RFC822 addresses into and out of this format. It would be easier,
and
> safer, to use RFC822 addresses in as literal a form as possible, so
we
> could just refer to that document and its descendants.

I agree completely. Tom's format was designed to look as like
as possible to http URLs, but there is in the
news:artic@leid a precedent for exactly the same reasons for
keeping the original as uncorrupted as possible.

> --john
>

Tim