[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: RE: Messaging Service CPP/CPA [was: Ack Message Payload??]
Hi, Thanks for the support. This question came up during the RM discussion. The use case was about resetting the sequence numbers. Either we can send a message with -1 to say, we are resetting the seq numbers or just roll over to 1. Both assume some implied intelligence and assumptions on the part of the message handler, when it handles normal (meaning normal, ack and error types) messages. But this shouldn't be, a standard should avoid any implied assumptions and all implications should be explicit ! Hence the logic for a new control message type. The reason why we didn't encounter this might have been because, only now we are incorporating RM and security. IMHO, security would need the control messages to sort out things between the message handlers. On further thought, may be the error and ack messages are sub-types of control messages. So may be the classification should be normal and control, with the control having a type attribute which can take multiple values (ack, error, RMControl, SecurityControl) et al. Of course, we can have the control messages as headers (like the RMHeader and SecurityHeader) and then embed the control directives in those tags. Reading your e-mail, I also realize that we might need a control message type for BP as well. cheers and have a nice day -----Original Message----- From: Nikola Stojanovic [mailto:nhomest1@twcny.rr.com] Sent: Sunday, October 29, 2000 10:12 AM To: Ebxml-Poc (E-mail); Ebxml-Transport (E-mail) Cc: Martin W Sachs/Watson/IBM Subject: Re: Messaging Service CPP/CPA [was: Ack Message Payload??] <Krishna> Can we also consider adding a <controlMessage/> as well, which could be used for message level control stuff like reset sequence numbers et al ? This would enable the messaging systems to exchange TRP level controls - including dynamic security reconfiguration, reliable messaging reconfiguration et al. </Krishna> Looks to me like a potentially valuable topic for CPP/CPA (TPP/TPA) -> "MS/Transport" capabilities (see TA draft, chapter 16) as well. However, I would suggest that representative Usages (UseCases) are needed before making any decision. Regards, Nikola
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC