I would guess that either most IdPs get it right or most SPs don't check. Also, implementers probably wonder (as I guess you did) what the difference is between Assertion IssueInstant and AuthnStatement AuthnInstant.
::Ari
> -----Original Message-----
> From: Tom Scavo [mailto:]
> Sent: Thursday, February 07, 2008 8:23 PM
> To: Eve Maler
> Cc: SAML Developers
> Subject: Re: [saml-dev] the value of AuthnInstant
>
>
> Thanks, Eve. I wasn't sure.
>
> Thinking out loud, I wonder how many implementations get this right?
>
> Tom
>
> On Feb 7, 2008 7:48 PM, Eve Maler <> wrote:
> > It's t1, isn't it? One assertion, issued at time t2, and another,
> > issued at time t4, will both indicate that the user authenticated at
> > time t1 -- assuming that authn session is still good, of
> course (e.g.,
> > re-authn isn't being forced).
> >
> > Eve
> >
> >
> > On Feb 7, 2008, at 4:04 PM, Tom Scavo wrote:
> >
> > > Suppose a user presents an AuthnRequest to an IdP at time
> t0. Since
> > > the user has no security context initially, the IdP challenges the
> > > user to authenticate, which the user does successfully
> (at time t1).
> > > So the IdP issues an assertion (at time t2), which the
> user transmits
> > > to the SP via the browser.
> > >
> > > At some later time t3, the user presents another
> AuthnRequest to the
> > > IdP. Since the user already has a security context, the
> IdP does not
> > > challenge the user to authenticate, but rather issues an
> assertion (at
> > > time t4), which again the user transmits to the SP.
> > >
> > > Question: What is the value of AuthnInstant in the second
> assertion?
> > >
> > > Thanks,
> > > Tom
> > >
> > >
> ---------------------------------------------------------------------
> > > To unsubscribe, e-mail:
> > > For additional commands, e-mail:
>
> > >
> >
> > Eve Maler +1 425 947 4522
> > Principal Engineer eve.maler @ sun.com
> > Business Alliances group Sun Microsystems, Inc.
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:
>
>
