In chapter "13.3.2 Activity", on page 429, there is the following bullet point:

"* After all completion dependencies have been fulfilled, the state of the Activity changes to Completed. The outgoing
Sequence Flows becomes active and a number of tokens, indicated by the attribute CompletionQuantity, is
placed on it. If there is more than one outbound Sequence Flows for an Activity, it behaves like an implicit
Parallel Gateway."

The last cited sentence does not take into consideration that the outgoing sequence flows might have conditions. Therefore, the statement "behaves like an implicit Parallel Gateway" is not entirely correct.

Suggestion for a revised version of the last sentence:
If there is more than one outbound Sequence Flows for an Activity, and if all outbound sequence flows are unconditional, the Activity behaves like an implicit Parallel Gateway. If there conditional outgoing sequence flows, the behavior is as described in "13.3.1 Sequence Flow Considerations".

Figure 10.39 states that Arbitrary Cycle is known as Workflow Pattern #16. This is not correct. Arbitrary cycle is Workflow Control Pattern # 10 in Van Der Aalst's documents. Recommend the 16 in the Figure description be changed to a 10.

Figures 10.18 and 10.19 are presented as though they are logical equivalents in the description above 10.19. However, In Figure 10.19 it is possible both Condition 1 and 2 could be true causing all three transitions to be traversed. This is not possible in 10.18 because of the exclusive gate. To make the two diagrams behave the same, the default slash should be added to the Condition 1 transition coming from the Inclusive gate on diagram 10.19. This will insure either Condition 1 XOR Condition 2 is traversed, but not both inclusively as is possible in the current spec. If the diagrams are not to be taken as logical equivalents, the text just under Figure 10.18 should be changed to make this clearer and the Activity names should be made different so equivalence is not implied.

I don?t know if this has come up in the BPMN RTF already? apologies if old news. TThe DMN beta followed BPMN?s external reference scheme of a prefixed ID value, but in the FTF version they changed to it to something like href=?[DMN filepath]#[ID]?. They say that the BPMN way is ?flawed? because it could lead to accidental ID collisions.?? I don?t understand why referencing a physical storage location in lieu of a proper namespace is less ???flawed? but my question is whether BPMN is moving to do it the DMN way.?? Seeing as DMN has specific pointers to BPMN elements, and I would expect some future BPMN to have specific pointers to DMN elements, I don???t see how the two standards could have totally different methods of external reference.

The Description for LoopCondition does not seem to be correct in "..., plus the timing when the expression SHALL be evaluated". The timing seems to be determined by the TestTime attribtue. If that is true, the "...plus ..."phrase should be taken out.

For the moment there is not MetaModel for BPMN. For MetaModel I mean a class diagram that shows the elements of the language and their relationships. In the paper Birgit Korherr, Beate List: Extending the EPC and the BPMN with Business Process Goals and Performance Measures. ICEIS (3) 2007: 287-294 authors give one proposal. It is very important to provide an OFFICIAL MM for BPMN, since it will bring clarity to the language.

In Figure 11.3, the annotation for the "repeat" indicator currently reads "The Sub-Process will repeat of the Discussion Over variable is False." It should read "The Sub-Process will repeat if the Discussion Over variable is False." The word "of" should be replaced with "if".

Hello, after reading throuh the named chapter about Sub-processes, I could not find a clear statement on the allowed number of "None Start Events" and "None End Events". Out of the spec one might have the feeling that a Sub-process always have a dedicated entry point from the "outside", which entry point is represented by a "None Start Event", and the sam epplies to End events/exit points. I could not find a statement regarding this, nor any overall principle out of which it could be concluded. It would maybe make sense to enrich the specification with this information.

the issue I'm reporting is a typo only: Chapter 9.4.2.3 contains information regarding reusable sub processes. The initial sentence introducing this chapter is as follows: <sentence> A Reusable Sub-Process object is an activity within a Process that “calls” to another Process that exists within a BDP (see Figure 9.10). </sentence> >From my point of view the listed abbreviation BDP is misspelled, I would expect BDP for Business Process Diagram. It's really a minor issue, but anyway I would like to point to

With regard to Timer Events, Expression in B.11.8 doesn’t provide a solution to the example given in the specification such as in Table 9.8 for Timer Events which use TimeDateExpression B.11.18 which in turn use Expression.

BPEL has three constructs For, Until and RepeatEvery. RepeatEvery can optionally be applied to the other two. The XSD excerpt is as follows:
<xsd:element name="for" type="tDuration-expr" />
<xsd:element name="until" type="tDeadline-expr" />
<xsd:element name="repeatEvery" type="tDuration-expr" />

Both the types of expressions extend tExpression which is defined as this:
<xsd:complexType name="tExpression" mixed="true">
<xsd:sequence>
<xsd:any processContents="lax"
minOccurs="0" maxOccurs="unbounded" />
</xsd:sequence>
<xsd:attribute name="expressionLanguage"
type="xsd-derived:anyURI" />
<xsd:attribute name="opaque"
type="xsd-derived:tOpaqueBoolean" />
<xsd:anyAttribute namespace="##other" processContents="lax" />
</xsd:complexType>
And are further qualified in section 8.3 as:
.Deadline expressions should return valid values of xsd:date and xsd:dateTime
.Duration expressions should return valid values of xsd:duration

We feel the BPMN spec is imprecise in this area in defining both in Table A.9 by their mapping to BPEL (TimeDate = until, TimeCycle = for). RepeatEvery makes no appearance in the BPMN spec.

Therefore, we think the best solution would be for BPMN to add RepeatEvery. Is it possible that the BPMN spec may have believed TimeCycle actually fulfils the BPEL repeatEvery, the name would seem to bear that out? However it explicitly says that TimeCycle should be interpreted as BPEL 'for'. Therefore a second, larger, change to BPMN would be to re-map TimeCycle to repeatEvery and add instead WaitFor or some such mechanism.

In BPMN Formal/08-01-17, Section 9.4.3.8, Reference Task, the first paragraph refers to Activity, while the second paragraph and the associated Table 9.31 refer to Task. For clarity and correctness, the first paragraph should also refer to Task.

To fix: Change the word "activity" to "Task" in the first sentence. Change the word "activities" to "Tasks" in the second sentence.

In the specs, it is stated that a gate can be designated as the 'Default' Gate' for Excusive Data-based gateway and Inclusive Gateway. It is stated in the specs for Gates that
"For DefaultGates: The Sequence Flow MUST have its Condition attribute set to Otherwise"
However, The Condition Type attribute for a Sequnce Flow can only be
Expression
None
Default

Should the specs state that "For DefaultGates: The Sequence Flow MUST have its Condition attribute set to Default"
Please confirm.

In the preface in the "Licenses" section it states "The companies listed above have granted to the Object Management Group, Inc. (OMG) a nonexclusive, royalty-free,paid up, worldwide license to copy and distribute this document". However other than the OMG itself no companies are listed above or in fact below in the preface. Surely this is bug - failing to list the copyrights of the companies that submitted the intellectual property. Note section 6.3 has a list of contributors. This is the URL I downloaded the PDF from: http://www.omg.org/spec/BPMN/1.1/ and the file was http://www.omg.org/docs/formal/08-01-17.pdf all linked from http://www.omg.org/technology/documents/br_pm_spec_catalog.htm