Re: [ebxml-bp] Signals and things - Q

["David RR Webber" <david@drrw.info>]

Dale,

 

OK - I see the issue - we do not need this to be a back-door thru which all kinds of stuff comes and goes.

 

A suggestion would be to make it unbounded - but add a annotation that explains that currently there

are 8 types and that we welcome feedback on additional ones, and that we reserve the right to limit

signal types in future once a field proven list has been determined.

 

DW

 

p.s. Brainstorming here - one another way to limit this is to suggest that signals be registered in a

      registry and that the BSI would check that specifically when processing the signal at the 

      messaging level - this would certainly limit proliferation of one-off signal types that degrade

      the interoperability value in the first place.

----- Original Message -----

From: Dale Moberg

To: David RR Webber ; BPSS ebXML

Sent: Tuesday, December 07, 2004 11:00 AM

Subject: RE: [ebxml-bp] Signals and things - Q

JJ suggested this value because the max of current signals for the Requesting and Responding Activities is 8 (RA, RAE, AA, AAE)

 

I guess this omits a mapping for a general failure signal, however (this had not been resolved at that time).

 

Unbounded would certainly meet any foreseeable signal proliferation that I can imagine, though.

 

I will be glad to adjust multiplicity to a TC consensus value though.

 

---

From: David RR Webber [mailto:david@drrw.info]
Sent: Tuesday, December 07, 2004 8:12 AM
To: BPSS ebXML
Subject: [ebxml-bp] Signals and things - Q

 

Team,

 

Just reviewing the materials to date ahead of todays call.

 

This is all coming together well - and the examples in

Appendix A and B definately very helpful.

 

Love figure 17 - super stuff.

 

Anyway - a dumb question - in the schema there is this

declaration:

 

<xsd:element name="SignalMap" minOccurs="0" maxOccurs="8">

 

Why 8?

 

Or was this just a typo - forgot to hold down the shift key while hitting the 8 key - wanting '*' for unbounded?

 

Thanks, DW