[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: [Moberg] 9/10/2004: Martin's Process Diagram and RosettaNet Dbl-Action
Given our conversation this morning about NOF, here is what general differences I see between the RosettaNet double-action [1] and Roberts' Business Activity Behavior. If I've missed something or changes are required to my interpretation, please advise. Two differences which we uncovered today (not those previously known) are shown with (***). * At Responder: o ***Differentiating the validation that occurs + RosettaNet: 'Validate Message Structure': As we discussed they have a message guideline. We didn't believe it was only syntactic validation. Is it cardinality, optionality, parsable and content, or which? + ebBP - Specific to Roberts' diagram - Syntactic the diagram indicates 'valid' and in sequence. o Acceptance Acknowledgment (or what it becomes) + RosettaNet: No longer uses. + ebBP: Shows processing on Responder for business request. * At Requester o Retries + ebBP - Specific to Roberts' diagram doesn't explicitly model the retries although they exist in the specification. + RosettaNet: Modeled the retry. o NOF + ***RosettaNet uses: # If TTP exceeded and a timeout occurs waiting on a Receipt Acknowledgment # If a timeout occurs and the retry count has been exhausted (i.e. retries = max retries) on a Receipt Acknowledgment # ***If they are unable to process a response in the backend system after a Receipt Acknowledgment has been sent. [2] + ebBP # Specific to Robert's diagram, doesn't show the backend system portion of processing. # No NOF currently exists. [1] Two-action activity/asynchronous [2] Similar to the case that Mukkamala mentioned. This is not timeout related.
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]