Computed ID Data Connector

[Cantor, Scott]

On 3/28/14, 10:19 AM, "Rod Widdowson" <rdw at steadingsoftware.com> wrote:

>Let me wind back and apologise.
>
>The sha is computed on the:
> - inboundMessageIssuer (i.e. theSP)
> - the input attribute
> and then salted.
>
>The IdP does entityID does *not* enter into the calculation.  I was
>getting
>my issuers and my responders and my relying parties confused.  My
>apologies

I thought it was hashed in myself, but I guess that's why it makes no
sense.

>Note that the CryptoTransient and Transient *attributes* (which are sort
>of
>kinda related except when they are not) *do* care about the IdP entityID.

They do in V2; I took that out of the V3 copies, I couldn't see any
compelling reason to keep doing that and at the time it seemed it might be
possible to narrow those APIs, which ultimately didn't happen anyway.

-- Scott