pem-dev
[Top] [All Lists]

Re: Linking authorizations to the DN

1993-08-17 15:55:00
Bob:

For that reason, I am not trying to cram authorizations into
the DN per se. Rather, I am trying to force a highly visible 
DISCLAIMER of any authorization into the digital certificate, 
which is the only place I can put in where my forger will be forced 
to use it. And unfortunately, I am caught in a standards box of
an X.509 certificate that doesn't accomodate what is needed,
and a PEM standard (and PEM implementations) that insist that
I have to use X.509, even though they don't define a minimal
or suggested set of attributes.

PERMISSIVE AUTHORIZATIONS ONLY WORK IF THE DEFAULT
IS NO AUTHORIZATION. Otherwiose they only add to your liability,
instead of subtracting from it. that is why we have to put a disclaimer 
in the X.509 certificate, not some other certificate, and the DN is the 
only place it can go.

I think the paragraphs above outline a basic philosophical difference in the
way in which we envision the use of PEM when supporting "security-enhanced"
applications such as EDI.  When used for simple email, there is no real need 
to add authorizations to a security-enhanced message.  All we really care
about is authentication (or for some, confidentiality).  Perhaps this is not 
true when the message purports to speak for an entire organization of which the 
individual is a member, but I will ignore this case.  When the messaging 
service is used as the vehicle for a larger application, such as EDI, we become
very concerned about authorization.  It has always been my opinion that when
supporting these applications PEM should provide nothing but authentication.  
Authorization should be provided through a separate mechanism such as the ECMA 
PAC, and that in the absence of explicit authorization, the default is none.  
In your terminology above, I would suggest a purely permisive model with a 
default of no authorizations.

It appears that you take the opposite view, that the PEM signature provides
authentication and absolute authorization, unless explicitly limited through
some mechanism such as a disclaimer in the DN.  But this is not how the 
paper world works, at least with most of the suppliers that I've dealt with.
Whenever I've opened up a line of credit with a supplier so we can use a 
simple paper PO, I've always been asked for a list of names authorized
to sign a PO.  This would fit into the permissive model.  In fact, I can't
think of any paper-world analogies in my working environment that do not
use the permissive model.  I wish it were the other way around.  I would
send the boss's secretary a memo granting myself a big raise, because I
doubt if he has explicitly prohibited my doing so.

Charlie Watt
SecureWare, Inc.

<Prev in Thread] Current Thread [Next in Thread>