RE: Why workflow is NOT just a Pi-process

Steve:

I thought I had acknowledged in my email that pi [being turing complete]
would allow you to model anything, ERP, workflow, ...

However in doing so it assumes that everything is a process, a variable, an
object, a GUI, a textfield, a user,...

If you wanted to model an ERP, I contend that the level of nesting (i.e. all
the composed processes from user to DB would be impractical to manage).
Except for Intalio which took a very pure approach to implementing pi,
everyone is using pi with one or two levels of composition (nesting). Which
I think is the wise approach.

A much bigger problem of pi is that you cannot efficiently define
"boundaries" if you start nesting/composing processes like this. If you put
yourself to the task of modeling a mid-size IT shop. Pi cannot be
efficiently leveraged to model ERP talking to CRM to SCM, ... It is not
impossible, rather my point is that pi itself does not provide you with the
tools to define these boundaries, they are at most a pattern or a discipline
of using pi. We end up in a rather strange situation, we wanted to use pi,
because it is a theory that works well for distributed system, but it works
well as long as you use a single process virtual machine !

Hence other efforts to carve these boundaries, such as BPELa or ws-cdl, and
the comment that choreographies are complementary. I go even further in my
last email, there is a duality between choreography and orchestration, as
proven by the fact that BPML 0.4 (an orchestration specification) was
repurposed with little change in a choreography specification (WSCI).

Steve, you might understand a bit my fustration, after 5 years, of
attempting to contribute to defining a business process models at BPSS, BPML
and WS-CHOR, I very sadly see that we are going pretty much nowhere,
completely ignoring the basics of what we are doing. I was at a meeting in
Paris last week, where one strong BPMI player was responding to my comment
that I could not understand why after 4 years of efforts we still did not
have a business process standard that had the concept of a "user" and
"organization" like WfMC had in 1995. His response was "errare humanum est"
and that was being fixed. How is it possible that someone designing a
"Business Process" specification forgets about the "user" like BPML and now
BPEL? Well it is possible, once you take an approach like Howard is taking.
Sadly, it is only a few individuals that decide what approach to take, in
one of the most undemocratic way.

You might call my approach syntactic sugar, I don't mind, at least I am
standing on the one that are trying to find a solution to the problem not to
match a solution to the problem. The syntactic sugar approach does not lead
to this kind of mistake.

Note that you made a rather ironic lapsus, because precisely pi is syntactic
and the sugar is semantic. 

So I reiterate my request: why are we spending all this time to create
specifications that are vastly incomplete and overlapping. Why can't we all
come to a point of discussion where we finally tackle this problem, creating
a solid infrastructure foundation (BPEL,ws-cdl,...) on which the "business"
semantics (i.e. the sugar) (BPML, ebBP...) can be built? I think answering
this question would prove very fruitful.

Ciao,

Jean-Jacques
tel: 425-649-6584
Cell: 508-333-7634

-----Original Message-----
From: Steve Ross-Talbot [mailto:steve@enigmatec.net] 
Sent: Wednesday, December 03, 2003 1:33 AM
To: Jean-Jacques Dubray
Cc: Greg Meredith; public-ws-chor@w3.org; 'Haugen Robert'
Subject: Re: Why workflow is NOT just a Pi-process


I really don't think that this debate is going anywhere.

I could say that everything can be represented as assembler but does that
help me in any way?
I could say that everything can be modeled in logic but does that help?

Alas saying that workflow is just a pi-process is not helpful.

The pi-calculus (and higher order forms of process algebra) can be used to
model many things, including workflow. After all the pi-calculus is turing
complete and has (as Greg Meredith has oft
stated) a number of interesting properties that allow us to reason
effectively about pi-calculus models.

I understand Howard's attraction to a simple unifying model of computation,
of which the pi-calculus is a candidate. So Howard is correct in asserting
that workflow is a pi-process but it's just not a useful thing to say for
most people.

On the other hand JJ suggests that you cannot represent workflow as a
pi-process because it lacks the higher level semantics (roles etc). Just
because it doesn't support them as syntactic sugar doesn't mean that you
cannot do so.

The fact that the pi-calculus unifies the notion of data and process is not
a bad thing. It is a good thing. It has a use in that you can model
interaction with data and with classical notions of process using the same
underlying model. This in turn allows you to reason about the systems that
you model more fully.

JJ also suggests that the pi-calculus works well for low levels of nesting.
I'm not quite sure I understand what is meant by "levels of nesting". But if
this is to include or be equivalent to recursive composition then the
pi-calculus is scale invariant with respect to composition. It doesn't
change and so is well suited to as many levels as you want to go to. So,
unless I misunderstand, the assertion is incorrect.

On the other hand JJ's point does beg the question "is it useful to model
workflow using the pi-calculus?.
To answer this (and I don't intend to at this point) we have to ask another
question, namely "what is it we want to know about a workflow that we can
only understand by leveraging the pi-calculus?".
If the answer is nothing then it is not useful to say "workflow is just a
pi-process". If there is something then clearly it is useful to say
"workflow is just a pi-process".

If it turns out that it is useful then we will be able to understand in what
context a workflow should be modeled in this way (is it just the execution
platform or the language that described the workflow?).

Perhaps better minds that mine could provide a single paragraph on why the
pi-calculus is useful in this regard.

Cheers

Steve T



On 3 Dec 2003, at 02:16, Jean-Jacques Dubray wrote:

> We may also want to keep Phil Wainewright in the loop, he is keeping 
> track of the debate and already wrote a very good summary:
>
> http://www.looselycoupled.com/blog/2003_11_23_lc.htm#10701540100113058
> 8
>
> This article could also be titled: "why pi does not matter..."
>
>
> 2. I think what we meant to imply by using the title "Workflow is just 
> a Pi process" is that there is something foundational about the Pi 
> concepts that allow us to model higher level processes, including 
> workflow-like processes.
> We could have equally written papers with the following
> titles:
>
> ERP is just a Pi process
> SCM is just a Pi process
> B2B is just a Pi process
> Adding Up is just a Pi process
> Managing A List is just a Pi process
> EAI is just a Pi process
> Data is just a Pi process
> etc
>
> <JJ>I think we should define what "models" means. If it means express 
> all the semantics of the things listed above as a pi-process, 
> Intalio's N3 designer tool might be falling a bit short.
>
> For instance, I understand that a PO business object can be modeled as 
> a process, actually it will require that all instance variable of this 
> object be a pi-process itself, not to mention the database in which it 
> is stored, and so on...
>
> I am wondering why SAP and PeopleSoft have not yet released a 
> pi-version of their ERP system?
>
> </JJ>
>
> Just to clarify. If you look at some of the swimlane diagrams in the 
> paper, each swimlane is a BPML process in its own right (the XML form 
> and notational form being just alternate notations).
> <JJ>I think you just touched the fundamental issue I have with pi, 
> BPML/BPEL and Intalio's product
>
> As I understand it, Pi and Intalio's product do not allow for role 
> separation (since a role IS A process). Since everything is a process, 
> there is no way to introduce the notion of domain of control or 
> independent role, everything is capable of exchanging messaging with 
> everything at all levels (that looks pretty evil to me, though Howard 
> seems to argue that this is great in the paragraphs below),
>
> I can easily conceive that choreography is not not needed as long as 
> you have process composition, but what choreography buys you is 
> precisely this notion of domain of control that does not seem to exist 
> in what I have seen so far. Pi works wonders in a wireless system 
> where the levels of nesting of pi-processes are relatively low. 
> However, I cannot imagine the disaster for an ERP system or a workflow 
> where the level of nesting could reach infinity (in computers terms).
>
> This is precisely because this notion of domain of control does not 
> exist that BPEL/BPML are going through excruciating pain to create the 
> concept of an abstract process, to precisely define the boundaries 
> that do no exist in the theory. As a matter of fact, pi just by 
> itself, cannot cross company boundaries. I think that this is what 
> howards says in the next two paragraphs.
>
> Maybe there is a business opportunity to run Intalio's Process Virtual 
> Machine in an ASP model, running all pi-processes in the world. Then 
> the problem will be solved.
>
> </JJ>
>
>
> The process virtual machine within the BPMS creates end to end 
> processes out of piecemeal processes at all levels. This is where the 
> Pi concepts come in, since the interaction between swimlanes is of 
> course mobile behavior as defined by Milner. We chose email as an 
> example as it is a recusive process with this characteristic. We have 
> found similar characteristics with change management processes, record 
> keeping processes etc. We see no correspondence between these and 
> typical workflow processes as WFMS have been typically applied in 
> business.
> It feels very different to me in practice.
>
> I believe that the significance of choreography lies at the heart of 
> this, which is kind of why a subset of BPMI.org members submitted 
> WSCI, based on BPML, to this group, as a first step towards 
> unification. After all, it is WSCI-territory that allows multiple 
> technologies from existing players (even if they have no intention to 
> build a BPMS) to be used in conjunction with each other.
>
> Howard

This email is confidential and may be protected by legal privilege. If you
are not the intended recipient,  please do not copy or disclose its content
but  delete the email and contact the sender immediately. Whilst we run
antivirus software on all internet emails we are not liable for any loss or
damage. The recipient is advised to run their own antivirus software.