RE: [xdi] reserved words disallowed for e-names

["Drummond Reed" <drummond.reed@cordance.net>]

Two quick responses to Loren, and then a third larger point around where
this discussion should really be going on.

1) First, Loren's definitions of "e-name" and "e-number" are IMHO correct.
The growing consensus is simply that "e-name" is a shorthand way of saying
"reassignable XRI", and "e-number" is shorthand for "persistent XRI" (we may
even suggest these terms be added to the XRI 1.1 spec). Therefore "global"
is necessary as a qualifier if one wants to refer to an absolute e-name or
e-number. However it's important to point out that even though "global
e-name" or "global e-number" will USUALLY mean oen assigned by an XRI
authority (i.e., =, @, +, or $), technically it also includes absolute XRIs
of other kinds, i.e., those that begin with URI authorities (//) or
cross-references (anything in parens).

2) I agree that the issue of reserved global e-names or e-numbers in the =,
@, or + spaces is therefore only an issue for those authorities, and not the
XDI TC.

3) Therefore the XRI/XDI list (see
http://idcommons.net/cgi-bin/mailman/listinfo/xrixdi) is really the place
for this discussion, and not the OASIS XRI or XDI TC lists.

I'm copying that list so we can continue the discussion there.

=Drummond 

-----Original Message-----
From: Loren West [mailto:loren.west@epok.net] 
Sent: Friday, May 14, 2004 12:34 PM
To: 'Fen Labalme'
Cc: 'XDI TC'
Subject: RE: [xdi] reserved words disallowed for e-names

My resistance may come down to a matter of definition.

I've always considered e-names and e-numbers as fancy
words for non-persistent and persistent XRIs (respectively)
that can be used anywhere in the XDI web.

One important place they're used is in the = and @
namespaces, and I've seen those e-names / e-numbers
referred to as "global e-names / e-numbers".

If, however, I've gotten my definition of e-name/e-number
wrong, and they really mean "the things registered under
the = and @ namespaces", then the usage of the term
"global" is redundant.

If e-names really are defined the way I thought, then
I have a problem with reserved words in the e-name
specification.  They should be in the "global e-name"
specification.

I understand the reason behind restricting single and double 
letter global e-names, but aside from "Example", I don't
see the reason to restrict any of those proposed e-names in 
the global e-name specification.

=Loren

-----Original Message-----
From: Fen Labalme [mailto:fen@idcommons.org] 
Sent: Friday, May 14, 2004 11:52 AM
To: Loren West
Cc: 'XDI TC'
Subject: Re: [xdi] reserved words disallowed for e-names


My thinking about why some names ought to be reserved regards 
documentation, similar to RFC2606 that reserves e.g. example.com.  Being 
able to write documentation that includes examples like 
'@community/member' and '=user' without these names referring to actual 
entities is, IMO, valuable.

I agree that if any names are reserved at all, it would be most 
important that they are reserved at the root, e.g., the = and @ name spaces.

Note that I also pose the question of disallowing one and two-character 
e-names, and e-names with any xri-reserved character in them (though the 
ABNF allows (e.g) this subset:  ;!*@&=+$,

Fen


Loren West wrote:
> I'm not sure why we need to reserve any of these names.
> If so, maybe only at the root.  Remember, XDI resources
> describe people (identities) as well as data within
> those identities (business cards, etc).
> 
> Although it doesn't make much sense to register people
> with those restricted words, I could imagine a few 
> thousand other words that don't make much sense, but
> I wouldn't know why we would want to disallow them.
> 
> Can you describe to me why you want to restrict e-names,
> and if so, in what namespace (=/$/@ - or some other namespace)?
> 
> Thanks,
> 
> =Loren
> 
> -----Original Message-----
> From: Fen Labalme [mailto:fen@idcommons.org] 
> Sent: Wednesday, May 12, 2004 10:55 PM
> To: XDI TC
> Subject: [xdi] reserved words disallowed for e-names
> 
> 
> At the TC meeting in New Orleans, I mentioned that we may want to 
> disallow some words as e-names in the same way the DNS Registries 
> disallow single letter names and (e.g.) example.com
> 
> I have taken a stab at an initial set of reserved words here:
>
http://xrixdi.idcommons.net/moin.cgi/EnamesAndEnumbers#head-bb873c830605d42e
> 40754efd118bfc9bcdb75c17
> 
> Please read and comment.
> 
> Thanks,
> Fen
> 
> PS: I'm not a member of the XRI list - perhaps this note should be 
> forwarded there, too?
>