As discussed on today's call - understanding the
issues and landscape
and how to position relative to where BPSS is right
now - see
Martins comments and my sense of possible short
term fixins -
although obviously from the BCM stance longer term
needs
beckon here...
DW.
----- Original Message -----
Sent: Monday, November 24, 2003 10:17 AM
Subject: Re: [ebxml-bp] XSD schema for OASIS BPSS
Martin,
There's two different but related issues here
IMHO.
First is simply the functionality in BPSS and the
coupling that can be
done.
Second is presenting this effectively to endusers
so that they can
manage what they need to manage.
I think we are on track to solve both.
Obviously for me - being able
to manage context is critical. Clearly that
is giving you all kinds of
grief in implementation - because there's only very
limited ways
to do this right now
in BPSS. We've opened an issue on this
already.
The second comes down to visualization
support. I realized when
I did the OAG AutoTech scenarios - that BPSS is
good for the
big picture - but when you need to dynamically draw
linkages
and flows - it has less support in the XML right
now. You can
fudge it - but for endusers - you need to be
specific explicit
constraints.
Monica - I think I need to open an issue and lead
it for this:
"Enhanced support for explicit flow pattern logic
and visualization" -
right now this is
implied - but we need more to cover off user
interfaces being able to express it. The
bottom line is that UML was
intended to be the answer here - but its not - and
so now we need
to add some semantics to the BPSS XML to make sure
diagramming
tools have all that is needed.
Thanks, DW.
----- Original Message -----
Sent: Monday, November 24, 2003 5:14
AM
Subject: RE: [ebxml-bp] XSD schema for
OASIS BPSS
Dear
all,
I have not been taking part in the calls as I have other
commitments that mean that 9pm calls are just not a good one for me.
However I though you might like to get some feedback on how our use of the
BPSS has been going.
We have had some interesting developments that
have changed our response to the BPSS.
1) we found that customers do not
seem to like to work with the activity diagrams. Even though the ones we put
out were accurate, someone else was asked to explain them in more detail and
subsequently issued a document with scenarios in that were far more
restrictive that they should have been.
2) we find that the state for the
Gateway is actually managed in most part by a CRM system. This has two
problems,
a) the state models implemented in
these systems is often not geared to a neat public/private split and therefore
the CRM user feel free to move business objects to states even if these are
not supposed to happen. (some of this is because we have automation robots
hook of the CRM platform and these need to have states set before they run.
Any failures mean that the states have to be reset and this is reflected to
the customer!)
b) The steps for the gateway need to
be easily identifiable, and this is proving very difficult as we find that
development issues creep in that mean that the main headline state is not the
only item being monitored and this leads to lots of confusion.
c) the need to be able to cope in the
BPSS with transactions being performed through other channels. This is a
really major issue. The implications are that the state of one partners system
can get out of step legitimately. When modeling this you find that you need to
run this transaction over the gateway in reverse. For example, customer places
order via phone. This would normally have come over the gateway, so the
customers System may now does not reflect this. So instead of order coming
from customer, a transaction needs to be enacted that get the customers order
back on to their systems.
The consequences of some of this are
that we have ended up resorting to simple message in and outbox type
transactions on our gateway and using the BPSS only as a documentation tool
expressing the ideal process.
With item c) above I think this shoul
dbe addressed by the team as without it you are saying that the BPSS only
really works in a pure environment which is not very reflective of real
life.
Martin
Roberts xml designer, BT Exact e-mail:
martin.me.roberts@bt.com tel: +44(0) 1473
609785 clickdial fax: +44(0) 1473 609834 Intranet Site :http://twiki.btlabs.bt.co.uk/twiki
Monica / Dale,
Where are we in planning to publish a draft
schema?
I'd very much like to have a new schema
available to the team
internally at least from Kavi - for testing and
development purposes.
I think this will also help alot with issue
diagnosis et al.
Thoughts on a plan? Can we discuss this
on today's call?
Thanks,
DW
|