-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

agreed .. I'm +1 for renaming those two composite assertions ..

Sanka


Paul Nolan wrote:
> Hi Sanka,
> This issue is closely related to the patch I sent WSCOMMONS-19.
> There is no
> doubt that an All can be modelled with a Boolean AND. However the
> ExactlyOne
> is not so obvious. It only behaves like an XOR for the case with two
> terms.
> If you bring up the truth table for XOR with three terms you will
> see that
> there is a row (first one) that does not have ExactlyOne term.
>
>
> A B C A^B^C
> ---------------
> T T T - T
> T T F - F
> T F T - F
> T F F - T
> F T T - F
> F T F - T
> F F T - T
> F F F  - F
>
> On 05/05/06, Sanka Samaranayke <[EMAIL PROTECTED]> wrote:
>
>>
>> Brian
>>
>>
>> Brian Hulse wrote:
>> > Dmitry,
>> >            I believe that Neethi is correct here because the
>> > specification
>> > states that "wsp:Policy is equivalent to wsp:All." This is with
>> > respect to
>> > its action upon its children but Policy does have other attributes
>> > such as
>> > Id, Name and extensible attributes. However, once again I have to
>> > agree with
>> > you ... AndComposite is not revealing or obvious, since it
>> corresponds
>> > to an
>> > All operator, so why not call it that?  I think that Neethi has
>> made an
>> > incorrect assumption that All behaves like an AND and ExactlyOne
>> behaves
>>
>> > like an XOR; this is plainly incorrect.
>> Well, Neethi has made the assertions All behaves like an AND and
>> ExactlyOne behave like an XOR.
>>
>> The All policy operator requires that *all* of its child assertions
>> should be met. IMO it has the similar logic to an AND which require
>> all
>> its inputs to be 1's to get 1 as the output. On the other hand
>> ExactlyOne requires that *only one* of its child assertion should be
>> met. In other words only one of its child assertions should be picked
>> for the final outcome (i.e. to load client side configurations).
>> IMO it
>> has the similar logic to an XOR which requires  exactly one of its
>> inputs to be 1 for the output to be 1.
>>
>> If you think these assumptions are incorrect, please explain so
>> that we
>> could consider to rename those composite assertions.
>>
>>
>> > ... I believe that Paul Nolan will be
>> > raising a JIRA issue on this shortly.
>> >
>> > Brian.
>> > On 5/4/06, Dmitry Goldenberg < [EMAIL PROTECTED]>
>> wrote:
>> >>
>> >> Hello,
>> >>
>> >> I have a question on the Neethi 1.0.1 API.
>> >>
>> >> What exactly is the difference between Policy and
>> AndCompositePolicy?
>> >> The
>> >> javadoc for both says that they require that all of their terms are
>> >> met.  I
>> >> assume the Neethi Policy corresponds to Policy in the WSP spec, and
>> >> AndCompositePolicy corresponds to Policy with the "All" alternative
>> >> type.  What exactly are the differences between the two?
>> >>
>> >> Thanks,
>> >> - Dmitry
>> >>
>> >>
>> >
>>
>>
>
>
> --
> Regards
> Paul Nolan
>

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.1 (GNU/Linux)

iD8DBQFEYCI6/Hd0ETKdgNIRAhyxAJ426xMEEOMbsr+P0LN+hf/2Nc5MjgCbB5NU
Hx4Ad+QJM9zapwm3zRAirYs=
=Grly
-----END PGP SIGNATURE-----

Reply via email to