RE: [ebxml-iic-msg] Re: follow-up on discussions

[Jacques Durand <JDurand@fsw.fujitsu.com>]

Title: RE: [ebxml-iic-msg] Re: follow-up on discussions

Matt:

We don't have to do tracing here - except that might help for debug.
The two Actions where I mention tracing, "Error" and "Mute",
as I describe them, won't give us much indication to help debug
in case anything goes wrong. However, we could modify them to get these
a little more verbose, message-wise:

- "Mute" is used as a dummy destination for messages that are intended
to cause some side effect in the MSH, like errors (so these mesg are not even supposed
to reach the app.)  Here a trace can tell if after all the MSH passed this
message to the app... even if it generates an error. But now another way to know this
is to get "Mute" send back a pre-canned mesg notifying that it was called, so that we
can process this on test driver side (I would not reuse the action "Responder" or "Initiator"
for this, as they rely on material from received message, which is precisely
supposed to be messed-up here).

- "Error" is an action that the MSH is supposed to call for "app-level" errors.
(I know only of two instances of that: (1) failure to receive Ack after all retries (see 6.5.4)
and (2) an error of severity level "Error" *should* notify the app of the MSH (see 4.2.3.2.4).
Here again a trace will tell us about this - but would be a manual process.
I actually have modified its behavior since last week in the doc I'm drafting, so that
"Error" action sends back a notice message to test driver.

I see tracing on the "other" side, as just a way to help figure what went wrong (has app
not been called? or has it but the returned mesg gone bad?),
but certainly not critical to the test case definitions and processing.
Opinion?

regards,

jacques

-----Original Message-----
From: Matthew MacKenzie [mailto:matt@xmlglobal.com]
Sent: Tuesday, May 28, 2002 10:10 AM
To: ebxml-iic-msg@lists.oasis-open.org
Subject: [ebxml-iic-msg] Re: follow-up on discussions

Jacques,

Please elaborate on your points 4.(c),(d).  Given that "tracing" is a
deprecated concept since our recent meetings, I do not fully understand
the rationale for these actions.

Thanks,

Matt

On Wednesday, May 22, 2002, at 12:03  AM, Jacques Durand wrote:

> Guys:
>  
> Here is my recap on recent discussions. Please review.
> I tried to describe (informally) some test cases using these framework
> assumtpions.
> Seems to work.
>  
> regards,
>  
> Jacques
>
>  
>
>
--
Matthew MacKenzie
XML Global R&D
PGP Key available upon request.

----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.oasis-open.org/ob/adm.pl>