Re: ISSUE 3564: Optional Assertions may not be usable in all circumstances

Prasad Yendluri wrote:
>
> One more thing. The spec clearly states that “an assertion with 
> optional=true” is syntactic simplification to having two policy 
> alternatives, one with and one without the assertion.
>

Disregarding the semantics for a moment, I've never been entirely clear 
what the benefit is of "normalizing" an optional assertion into two 
policy alternatives? As far as I can see, the same goal could be 
achieved by these rules:

- The intersection of an assertion with optional=true with the same 
assertion with optional=false yields the assertion with optional=false.
- The merge of an assertion with optional=true with the same assertion 
with optional=false yields the assertion with optional=true.

Fabian

-- 
Fabian Ritzmann
Sun Microsystems, Inc.
Stella Business Park             Phone +358-9-525 562 96
Lars Sonckin kaari 12            Fax   +358-9-525 562 52
02600 Espoo                      Email Fabian.Ritzmann@Sun.COM
Finland

Received on Friday, 29 September 2006 09:16:57 UTC