I think I would need to see a clearer example, but I don't believe that what
you are suggesting is very far from what I  suggested so it should work.
What I suggested for item 3 could easily be done the first time it is used
in a section and then use the "normal" term from that point on.  The clearer
definition would amount to the table that I suggested.  I don't believe that
we have any locations currently where a single technology is used more than
once and between more than two entities.

 

From: [email protected] [mailto:[email protected]] On Behalf Of
Tschofenig, Hannes (NSN - FI/Espoo)
Sent: Saturday, December 17, 2011 8:32 AM
To: [email protected]
Subject: [abfab] Issue 7 - Entity Naming Problems Galore

 

Hi Jim, 

in issue #7 you raise the following issue:

 <http://trac.tools.ietf.org/wg/abfab/trac/ticket/7>
http://trac.tools.ietf.org/wg/abfab/trac/ticket/7

"

There are massive problems throughout the document in that we are not using
a consistant set of names for each of the entities in the document. Part of
this issue is that there are different names for each of these entities in
each of the different protocols that we are using. However it also makes the
entire document difficult to follow as one needs to keep track of this all
of the time. 

THe following things need to be done:

1.      A table that maps from the name used in the document to the name
used in each of the different protocols 

2.      The use of a consistant name for each entity 

3.      Where feasible, you can do something like RP (Service Provider) so
that both names are in the text. 

"

I agree with the challenge regarding the identity terminology in general and
the problem of how the existing terminology used in various protocols aligns
with the other protocols and the overall framework. Your suggestion will be
difficult to implement, I believe, if you think about the figures and the
message flows. A direct mapping isn't easy either. 

In fact what we do most of the time in the document is use abstract concepts
(like RP, IdP) and then instantiate them for our use case. In our
architecture the IdP is using the AAA protocols (and as part of it EAP).  

I am curious whether we could get away with better defining the abstract
terms and in the other parts of the document where it matters focus on the
technology only. For example, when we talk about the relying party we would
that for it hosts the AAA client. 

Does this make any sense to you?

Ciao
Hannes

_______________________________________________
abfab mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/abfab

Reply via email to