Re: Issue Group 5 BALLOT (revised)

[Jeff Hodges <jhodges@oblix.com>]

(sigh -- got bit again by the ol' word-wapping (or not) bugaboo. the file
attached has "hard" line wrapping)

My ballot on this is attached. 

I have my positions on Group1 and Group3 figured out, but am out of time to
write them up today, and so will be doing that tomorrow. I hope this is ok. 

I have comments on the Group5 ballot itself that I'll write up tomorrow also. 

thanks,

JeffH
ISSUE:[UC-5-03:AuthCThrough]

(a) Resolution: Yes/No

No.

rationale: It seems to me that SAML is effectively a
meta-authentication protocol. As such, it should make use of (i.e.
leverage) the plethoria of extant authn mechanisms, and it
shouldn't attempt to expressly be one itself (for some definition
of "being one"). 


(b) Resolution: Yes/No

Yes. But the rationale should be clearly explained. 


(c) Resolutiom: Yes/No

No.

rationale: the suggested phrasing of the [NO-AuthC] "non-goal" in
the group5-ballot itself needs reworking. It is vague as presently
stated.

-------------------------------------------------------------------
ISSUE:[UC-5-02:SASL]

Resolution: Yes/No

[SASL] RFC 2222:

Yes.

rationale: it may or may not be appropriate to mention SASL in the
use case & reqs phase (and document(s)), but I do feel it will
definitely be appropriate to consider it in the context of the
-protocol and/or -bindings facets of SAML; hence we shouldn't
specifically exclude it, and thus the wording of this issue
[UC-5-02:SASL] in the group5-ballot is appropriate in terms of a
general resolution of intent, but whether and how we put it in any
-use work products is yet another question, imho.

---------------------------------------------------------------------

[ISSUE:[UC-5-01:AuthCProtocol] 

(a) Resolution: Yes/No

Yes.

rationale: I support the removal of that particularly worded
statement because I feel it is vague, and in consideration of my
responses to [UC-5-03:AuthCThrough] and [UC-5-02:SASL] above.

(b) Resolution: Yes/No

Yes.

rationale: "yes" is the short answer. the longer answer includes
some refinement of the wording of both [R-StandardCreds] and
[R-ExtensibleCreds] that I'll suggest in comments upon the
Group5.txt write-up itself.