Re: [certid] Need to define "most specific RDN"

[Peter Sylvester <[e-mail]>]

I am not sure that the discussion about constraints is
the most pressing thing we need to talk about. ;-)

Anyway, as Martin already summarized, the usage
of common name to identify a server is deprecated
since the last millenium.

It is nncoding of a *complete* identifier as a subcomponent
of another one, something that can happen for
non technical purposes, e.g. user friendliness etc.

Building software based on such ad hoc approaches
requires pattern matching logic, etc, not very difficult
to implement but not necessarily easy to use.

In fact, I think that the whole of defining a thingy CN-ID
complicates the whole text, it should just be a small
note indicating that is is a particular encoding of exactly
one DNS-ID.

It seems to me that the changes in the last version already
go into this direction. Nevertheless, the last paragraph of
2.2 (starting with "We recognize") introduces (AFAIR) a
string representation of a DN which doesn't serve any
real purpose and which has a strange "hierarchy". Note,
that one explains that a DN in X.500 is a hierarchy.
The implementation note refers to what? I see some confusion
regarding the example.

Just because we have mentioned all kinds of adjectives during
the discussion, I do not see the value mentioning them. The
only one that matters is the the "(most specific)" mentioned
in RFC 2818 IMO. There are some people knowning X.500
quite well: If one respects the non-normative schema rules,
can one have more than one Common Name AVA in a DN?


It has also been said that using the reerm "representation" is
unfortunate. In corresponding texts (X.500 etc), the word
'indication'  is used. I suggest to change this all over in the
document.

/P