[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [wsbpel] Issue 154 - Proposal For Vote
Just to be clear: this is not a failure of BPEL, in fact this problem is
shared by WSDL 1.1 users as a whole. So using this issue for suggesting
again that the BPEL spec is broken is just nonsense. The fact is that there
are holes in the WS stack, but that doesn't mean that BPEL needs to take on
the job of patching them all to succeed.
Anyhow, I think enough has been said on such a minor issue. I would suggest
we take a vote and get on with it.
Paco
"Yaron Y. Goland"
<ygoland@bea.com> To: Francisco Curbera/Watson/IBM@IBMUS
cc: wsbpeltc <wsbpel@lists.oasis-open.org>
02/04/2005 01:53 Subject: Re: [wsbpel] Issue 154 - Proposal For Vote
PM
Please respond to
ygoland
One of the qualities of a good specification is that it doesn't make any
implicit assumptions. Currently the BPEL specification fails to meet
this bar as it implicitly assumes that 'somehow' the binding layer will
be able to produce a value that can be placed in a messageType variable
but it never makes this assumption clear in the spec.
The result is that there are circumstances in WSDL where a legal WSDL
cannot necessarily be unambiguously decomposed from a binding layer
message to a messageType. Because of the BPEL spec's failure to be clear
the implementer is left confused as to who bears the responsibility of
fixing the situation, the binding layer or the BPEL processor.
My proposal is that we make explicit what until now has been implicit -
if there are any issues translating between the binding layer and the
messageType then the responsibility for resolving matters belongs to the
binding layer and not BPEL.
Yaron
Francisco Curbera wrote:
> The text is certainly a little better but I still think we are not doing
> anyone a favor by sprinkling the BPEL spec with WSDL issues and warnings
-
> there are way too many of those! My preference is not to address this
issue
> in the BPEL spec.
>
> Paco
>
>
>
>
>
>
> "Yaron Y.
> Goland"
>
>
> <ygoland@bea.com> To: Francisco
> Curbera/Watson/IBM@IBMUS
>
> cc: wsbpeltc
> <wsbpel@lists.oasis-open.org>
>
> 01/14/2005 06:26 Subject: Re: [wsbpel]
Issue 154
> - Proposal For Vote
>
>
> PM
>
>
> Please respond
> to
>
>
>
> ygoland
>
>
>
>
>
>
>
>
> I agree with your statement that "In BPEL we need to assume that WSDL
> and SOAP processing have taken place and provide us with the right
> messages, properly processed.", what the issue is asking for is that the
> assumption be made explicit.
>
> Perhaps a warning of the form "BPEL assumes that the WSDL binding layer
> is able to decompose incoming messages into the parts specified by the
> WSDL message definition. However it is know that certain combinations of
> message definitions and bindings, including ones defined in the WSDL
> standard itself, cannot be decomposed in any standard way. For example,
> a multi-part WSDL message where one of the parts is a complexType and a
> doc/lit SOAP transport can create ambiguous situations. The BPEL
> specification assumes that these ambiguities will be dealt with at the
> binding layer, perhaps by forbidding ambiguous message definitions, and
> are therefore out of scope of BPEL."
>
> Yaron
>
> Francisco Curbera wrote:
> > I personally prefer to leave this type of WSDL or SOAP
recommendations
> to
> > the appropriate 'authorities' (WS-I and the like). In BPEL we need to
> > assume that WSDL and SOAP processing have taken place and provide us
with
> > the right messages, properly processed. The use case mentioned here
might
> > make sense for example when a non-SOAP serialization is encountered.
So
> my
> > preference is to close w/o change.
> >
> > If (against this advice) the TC persists in addressing this issue in
the
> > BPEL spec, I would suggest that the last sentence ("implementations
> > reject...") be changed to a recommendation to authors to ensure that
the
> > message definitions they use don't lead to processing difficulties
under
> > certain bindings. Based on BPEL's binding-independence aims this seems
> like
> > a more reasonable piece of advice to me.
> >
> > Paco
> >
> >
> >
> > |---------+---------------------------->
> > | | "Yaron Y. Goland"|
> > | | <ygoland@bea.com>|
> > | | |
> > | | 01/13/2005 08:47 |
> > | | PM |
> > | | Please respond to|
> > | | ygoland |
> > |---------+---------------------------->
> >
> >
>
>-------------------------------------------------------------------------------------------------------------------------------------|
>
> >
> >
> > |
>
> > |
> >
> > | To: wsbpeltc
> > <wsbpel@lists.oasis-open.org>
>
> > |
> >
> > |
> > cc:
>
> > |
> >
> > | Subject: [wsbpel] Issue 154 - Proposal For
> > Vote
> |
> >
> >
> >
>
>-------------------------------------------------------------------------------------------------------------------------------------|
>
> >
> >
> >
> >
> > Issue 154 - doc/lit & multiple body parts
> >
> > Proposal: Put in an implementer's note on the problem of multi-part
> > messages defined using complexTypes that are encoded using doc/lit
> >
> > Rationale: It is possible to create message definitions using doc/lit
> > and complex types where it is impossible to decompose the messages
into
> > parts. But this problem doesn't exist with rpc/encoded, just with
> > doc/lit. Since BPEL doesn't operate at that level we can't officially
> > ban the practice. But we can at least warn people.
> >
> > Changes Required:
> >
> > Section 3 -
> >
> > Insert new paragraph after the paragraph that begins "While WS-BPEL
> > attempts to provide as much compatibility with WSDL 1.1 as
possible..."
> >
> > WS-BPEL only operates at the WSDL portType layer and so intentionally
> > does not address WSDL binding issues. But there is one particular WSDL
> > binding issue that implementers should be aware of. If a WSDL message
is
> > defined with multiple parts at least one of which is defined using a
> > complex type and the resulting message is bound using doc/lit then it
is
> > at least theoretically possible to create a situation where it is
> > impossible to determine for a message instance where one part of the
> > message ends and another begins. WS-BPEL only requires that messages
be
> > broken into their required parts, it does not specify how, therefore
> > this issue is out of scope for WS-BPEL. But in general it is
recommended
> > that implementations reject message definitions where this ambiguity
> > exists. Note that WS-I's Basic Profile explicitly forbids WSDL
messages
> > constructed as described above.
> >
> > To unsubscribe from this mailing list (and be removed from the roster
of
> > the OASIS TC), go to
> >
>
http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php
>
> > .
> >
> >
> >
>
> To unsubscribe from this mailing list (and be removed from the roster of
> the OASIS TC), go to
>
http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php
> .
>
>
>
>
> To unsubscribe from this mailing list (and be removed from the roster of
the
> OASIS TC), go to
>
http://www.oasis-open.org/apps/org/workgroup/wsbpel/members/leave_workgroup.php
.
>
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]