Fw: Latest Editor's WS-CDL document dec-03-2004

----- Original Message ----- 
From: Gary Brown 
To: Nickolas Kavantzas ; member-ws-chor@w3.org 
Sent: Saturday, December 04, 2004 3:33 PM
Subject: Re: Latest Editor's WS-CDL document dec-03-2004


2.5.2 Interaction

p49, line 11:

"When two or more record elements are specified for the same Role in an Interaction with target Variables of Exception Type, one of the Exception recordings MAY occur. An Exception recording has an non-observable predicate condition, associated implicitly with it, that decides if an Exception occurs"

Not sure what this is trying to say - can anyone enlighten me? Sounds non-deterministic to me.

If the exhange or timeout associated with the record occurs, and the record has a 'causeException' set to true, then it should be deterministic and observable that the exception occurs. If causeException is false, then it should not cause an exception.

Regards

Gary

  ----- Original Message ----- 
  From: Gary Brown 
  To: Nickolas Kavantzas ; member-ws-chor@w3.org 
  Sent: Saturday, December 04, 2004 2:49 PM
  Subject: Re: Latest Editor's WS-CDL document dec-03-2004


  Comments:

  Section 2.4.2 Variables:

  Change: (silent attribute para)

  "....Choreography, if these operations should not be observable to other parties. The default value for this attribute is "false"."

  to:

  "....Choreography. A 'silent' variable is used to represent the result of activities within a party that are not observable (or relevant) from the CDL perspective. The default value for this attribute is "false".

  **********

  Change: last para beginning "The attribute name ..."

  Remove from "The Variables with Role Type not specified ..." up to "respect to the Role Type(s) in which the Variable resides.

  This is because these semantics are redundant now that role type list is supported, and having multiple variables with the same name would be confusing - especially if those variables had different types.



  *******

  Section 2.4.3.1 CDL Functions

  First function (getCurrentTime) is not on separate line

  ********

  hasDeadLine passed function definition is not in italics.

  Unclear semantics? Is it the intention that when these deadline/duration passes, that a blocked function will become unblocked even if all of the variables are not set? Or is the intention that the condition will continue to block, and only call these functions when all the variables are available, in which case it is not really acting like a timeout. Possibly would be better to model this on a different attribute on the workunit?

  No suggested text as I was not clear of the intended semantics.

  ***********

  Additional statement on globalizedTrigger function explanation:

  "Only one expression can be defined per role name."

  ********

  Section 2.4.4. Tokens

  First paragraph includes an example that implies that tokens can be used within conditions. This is not the case in CDL, where tokens are only used for reference and identity fields in a channel type. Would suggest removing from "For example...." to the end of the para.

  *********

  Change from:

  "alphanumeric and a counter an"

  to:

  'alphanumeric and a "Counter" an'

  *********

  2.4.5 Choreography

  Choreography Syntax: why/when was the exception and finalizer names changed to append Block? Don't remember this being part of the proposal that was voted on.

  *********

  Section 2.4.6 WorkUnits

  Page 32 (top paragraph)

  Nobember 11 Draft:

  "If the attribute block is set to "true" and one or more required Variable(s) are not available, then the Work Unit MUST block. When the required Variable information specified by the guard condition become available and the guard condition evaluates to "true", then the Work Unit is matched. ...."

  Latest Draft:

  "If the attribute block is set to "true" and one or more required Variable(s) are not available or the guard condition evaluates to "false", then the Work Unit MUST block. When the required Variable information specified by the guard condition become available and the guard condition evaluates to "true", then the Work Unit is matched.. ......"

  The slight difference, with a significant impact on the semantics, is the "or the guard condition evaluates to 'false'". This implies that activities that follow the workunit will only be performed if the workunit condition evaluates to true - otherwise it will block (potentially indefinitely).

  When was this change agreed? What issue number is it addressing?

  ***************

  Section 2.4.8 p34 second para:

  Additional sentence following:

  If the Exception Block is not present, the Choreography implicitly enters the Closed State.

  Additional part is:

  If the Exception Block is not present, the Choreography implicitly enters the Closed State and the exception is propagated to the enclosing choreography.

  This matches the semantics described in the exception rules (4th rule) on page 36.

  **********



    ----- Original Message ----- 
    From: Nickolas Kavantzas 
    To: member-ws-chor@w3.org 
    Sent: Saturday, December 04, 2004 1:21 AM
    Subject: Latest Editor's WS-CDL document dec-03-2004


    http://lists.w3.org/Archives/Public/www-archive/2004Dec/0004.html