The avoidance of omphaloskepsis is
definitely desirable.
The more important point IMO is whether or
not the state machine, which constitutes the protocol, implements the desired
service characteristics or not.
Since that rigor has not yet been
presented, the DA discussion is folly, unless DA(s) are intended to be a
description of the service characteristics that we are trying to achieve.
-bob
From: Christopher B
Ferris [mailto:chrisfer@us.ibm.com]
Sent: Thursday, October 20, 2005
3:08 PM
To: ws-rx@lists.oasis-open.org
Subject: Re: [ws-rx] Proposal for
AI 40 and i024
I don't get it.
Why
all the fuss over something that is purely informational? Why all the bother
over
the
use of the term "observed"? Why is it necessary to add something to
the mark-up
at
all? The definition of the parameter/assertion is what the creators of that
assertion/parameter
choose
it to mean. If we choose it to mean: this has no effect what-so-ever on the
manifestation
on-the-wire, then it has no effect on the manifestation on-the-wire.
Can
we stop the obsessive nonsense and move on to issues that are important?
As
I have maintained all along, and as I believe the TC members have come to
consensus
agreement
on, the DA is a contract between the RMD and AD. Period. How it is implemented
is
immaterial to the spec. We agreed to define an INFORMATIONAL
assertion/parameter
that
can be used to inform the RMS (or AS) as to what it can expect the RMD to do
with the
messages
it receives.
Whether
or not, there is a specific WS-Policy attribute that discriminates between
assertions
that manifest themselves on the wire or not is irrelevant IMO.
Frankly,
I would like to move on from this navel-gazing exercise. Do we REALLY think
that
someone
is going to be confused by this? I mean, really?
IMO,
whether we use "in effect" or "observed" matters little in
the end.
Cheers,
Christopher
Ferris
STSM, Emerging e-business Industry Architecture
email: chrisfer@us.ibm.com
blog: http://webpages.charter.net/chrisfer/blog.html
phone: +1 508 377 9295
Anish
Karmarkar <Anish.Karmarkar@oracle.com> wrote on 10/20/2005 02:08:03 PM:
> Here is my reason for agreeing to the
resolution at the F2F:
>
> There are two kinds of policies: policies
that affect the messages on
> the wire and policies that do not (or
"observed"). "observed" policies
> would be ones like privacy policy of a Web
site or auditing. These kind
> of policies are informational to the client
of the service and the
> messages on the wire do not get affected by
it. Non-observed policies
> are policies that affect the messages. For
example, encryption etc.
>
> My sense of the F2F resolution was that the
TC wanted to capture the
> fact that DAs, timeouts etc were
"observed" and an AI was generated to
> capture this (the wsp:Observed in the old
ws-policy doc does not define
> this well).
>
> From that perspective, "in
effect" doesn't quite capture our intention.
> "observed" is tainted because of
historical baggage. How about just
> stating that the RM assertion parameters do
not affect the messages on
> the wire?
>
> Comments?
>
> -Anish
> --
>
>
> Marc Goodner wrote:
> > I realize that, the resolution to i024
was to close the issue and
> > clarify the meaning of the term
“observed”. The word “observed” was not
> > added to the spec by i024, it was
already there. My proposal drops the
> > ambiguous term and adds text that does
not need any clarification.
> >
> >
> >
> >
------------------------------------------------------------------------
> >
> > *From:* Ashok Malhotra
[mailto:ashok.malhotra@oracle.com]
> > *Sent:* Tuesday, October 18, 2005 12:50
PM
> > *To:* Marc Goodner;
ws-rx@lists.oasis-open.org
> > *Subject:* RE: [ws-rx] Proposal for AI
40 and i024
> >
> >
> >
> > The word 'observed' alone would not have
satisfied the issue.
> >
> > That's why the WG wanted words to
clarify the semantics.
> >
> > All the best, Ashok
> >
> >
> >
> >
> >
> >
------------------------------------------------------------------------
> >
> > *From:* Marc Goodner
[mailto:mgoodner@microsoft.com]
> > *Sent:* Tuesday, October
18, 2005 12:40 PM
> > *To:*
ashok.malhotra@oracle.com; ws-rx@lists.oasis-open.org
> > *Subject:* RE: [ws-rx]
Proposal for AI 40 and i024
> >
> > I don’t understand
why defining the term “observed” would have
> > satisfied this issue and
the change below does not. I see no
> > ambiguity in the words I
have proposed.
> >
> >
> >
> > What is ambiguous about
“in effect”? Is not the intention of the
> > definition of the DA here
to expose what DA is in effect between the
> > RMD and AD where this
would be used? I can see some ambiguity about
> > “observed” in
this context as it is a word that can have
> > unintentional
interpretations among WS-Policy experts. Particularly
> > for those who have
differing recollections of the meaning of
> > wsp:Observed before it was
removed from WS-Policy. I don’t believe
> > the word
“observed” was used here to cause such interpretation and I
> > thought that was why
clarification was needed.
> >
> >
> >
> > I would have expected the
clarification of “observed” to say
> > something about the DA
being “in effect” between the RM and app
> > layer. Removing the
offending word seems a cleaner approach than
> > attempting to define it.
> >
> >
> >
> > If the below change is
acceptable to the TC I see no reason to
> > reopen the issue.
> >
> >
> >
> >
------------------------------------------------------------------------
> >
> > *From:* Ashok Malhotra
[mailto:ashok.malhotra@oracle.com]
> > *Sent:* Tuesday, October
18, 2005 12:15 PM
> > *To:* Marc Goodner;
ws-rx@lists.oasis-open.org
> > *Subject:* RE: [ws-rx]
Proposal for AI 40 and i024
> >
> >
> >
> > So, the proposal is to
change the undefined word "observed" with
> > alternate undefined
> >
> > words "in
effect". I'm sorry, but I disagree.
> >
> >
> >
> > I would like to ask the
chairs to reopen the discussion to determine
> > what the WG really
> >
> > wants.
> >
> > All the best, Ashok
> >
> >
> >
> >
> >
> >
>
------------------------------------------------------------------------
> >
> > *From:* Marc
Goodner [mailto:mgoodner@microsoft.com]
> > *Sent:*
Tuesday, October 18, 2005 9:42 AM
> > *To:*
ws-rx@lists.oasis-open.org
> > *Subject:*
[ws-rx] Proposal for AI 40 and i024
> >
> > I propose
the following to close the open AI 40 for i024,
> > clarifying
the meaning of observed.
> >
> >
> >
> > Change the
word “observed” on lines 233 and 245 to “in effect”.
> >
> > Change the
words “observed by” on lines 233 and 245 to “in
> > effect
at”.
> >
> >
> >
> > This would
result in the following text in the RM Policy doc.
> >
> >
> >
> > The Delivery
Assurance indicates a delivery assurance claim *in
> > effect*
between an Application
> > Source and
an RM Source or an Application Destination and an RM
> > Destination.
> >
> >
/wsrmp:DeliveryAssertion
> > An assertion
that makes a claim as to the delivery assurance
> > policy *in
effect* *at* the
> > destination
endpoint.
> >
> >
> >
> > I believe
that this retains the intention of the current text.
> >
> >
> >
> > 1 AI 40
> >
http://www.oasis-open.org/apps/org/workgroup/ws-
> rx/members/action_item.php?action_item_id=1048
> >
> >
> > 2 i024
> >
http://www.oasis-open.org/apps/org/workgroup/ws-
>
rx/download.php/14894/ReliableMessagingIssues.xml#i024
> >
> >
> >
> >
>