[ietf-dkim] New Issue: The term identity in the overview
dhc at dcrocker.net
Wed Mar 26 13:11:01 PDT 2008
I fear that this is a real matter of technical confusion -- not just literary
taste -- and that the source of confusion is common to the topic and not just
how the document is written.
That's not to say that the writing might not be contributing to the confusion.
So let's see if we can have a brief dialogue to clarify the points of
distinction, which might make better wordsmithing easier.
The current wording is trying to emphasize the difference between reference and
referrant, or the thing and a label for the thing. Identity is meant to refer
to the thing itself, such as you, where as identifier is meant to refer to a
label, such as "Stephen Farrell".
From my own historical confusion in these sorts of discussions, as well as the
confusion i keep seeing in others, making and maintaining the distinction
between thing and label is very real challenge, often failed.
To the particulars of the document:
A domain name is an identifier. The organization owning it is the identity.
The trademarked name of the company is another identifier. Sometimes the
domain name is treated as the online name of the company, as well as its
"address". That's ok. It's still an identifier, even if it is thought of as
similar to the registered corporate name. (Distinguishing between domain name
and company name might be important for assessment services, but it isn't for
DKIM signature validation.)
So with the above as background, can you elaborate a bit on the confusion
you are seeing? Anything that you or others can provide might help us better
understand the underlying writing problem that we need to address.
Stephen Farrell wrote:
> (As a participant only again)
> The use of the term "identity" throughout is somewhat confusing. I
> don't have a simple change to suggest. As an example, the start of
> section 2 says: "Given the presence of that identifier, a receiver can
> make decisions about further handling of the message, based upon
> assessments of the identity that is associated with the identifier."
> I find that hard to understand and could imagine it being very confusing
> to general readers.
> My suggestion would just be to ask the editors to give it a pass
> where they check that the various uses of this and related terms
> are ok and clear. As far as I'm concerned this issue can be closed
> as soon as they say "yes, we did that"
> NOTE WELL: This list operates according to
More information about the ietf-dkim