The resolution to ESSENCE-166, adopted in Ballot 3 did not include the following part of the proposed solution given by the issue submitters:

Add new section “9.7.6.8.4 Competency Level Detail Card” similar to section “9.7.4.7.5 Alpha State Detail Card”.

Add subclause 9.7.7.8.4 as requested.

Problem to be solved: The fourth OCL expression on practices makes two illegal references to "cc.workProduct". In addition, there is a typo in a reference to "self.allEments".

Proposed solution: Replace both occurences of "cc.workProduct" by "cc.levelOfDetail" and correct the typo by writing "self.allElements".

Agreed.

Request a change be made to the phrase "completion criteria" under Kernel Activity Spaces, to "primary goals".

Rationale: Phrase "Completion criteria" is leading people to the wrong understanding, such as the wrong belief that activities will end when "completion criteria" met.

"primary goals" is a more accurate phrase for the intent.

Request a change to the symbol used for Activity Spaces from the current rectangle with a "pointie head" to a basic rectangle with NO pointie head similar to the Areas of Concern symbol.

Rationale: The symbol is causing confusion to people leading them to think this is an activity rather than a generic container for activities and leading them to believe their is an implied sequential order which is not the case.

Issue 176 has a proposed resolution to change the name of the Bounded Requirements alpha state to Scoped.

However there is still a need to change the definition to better reflect the checklist and the concept of scoping with respect to its making the extent of the system clear.

We need to have a high level definition which captures the concept of clear extent and purpose.

the current definition in 8.2.3.1 does not capture the extent concept of scoping or boundedness properly.

"Bounded: The purpose and theme of the new system are clear."

Also, if you look at clause 9.8.3 in the examples of the textual notation there is a another definition which is different than the one in the alpha section, but it uses the "extent" concept

"state Bounded

This is adds extent concept but keeps the "theme" concept.

The concept of "theme" seems too loose for a proper definition.

Proposed Resolution

I suggest that we resolve this issue by replacing the definition in
8.2.3.1 and in 9.2.3 for the new
named state "Scoped" with the following:

"Scoped: The purpose and extent of the new system are clear"

Updated the definition of the Bounded state as suggested (independent of the resolution to Issue 176). (Note, however, that the subclause references given in the issue description are incorrect. They should be 8.3.2.1 and 9.8.3.)

In Figure 154, the Lifecycle Diagram for "Multi-Phase Waterfall Requirements Alpha Extensions" the requirements state labeled "Described" should be labeled "Acceptable".

Attached is a corrected figure which has change the state "Described" to "Acceptable"

Agreed.

Problem to be solved: The spec allows Activities to perform four operations (create, read, update, delete) on a Work Product or an Alpha. The card just names input and output, which is less specific. Activity Spaces just allow input, which seems very limited. In general, the language spec seems unbalanced with respect to this aspect.

Proposed solution:

• Remove association “input” from language element “ActivitySpace”.
• Remove language element “Action” and all its associations.
• Add associations “input” and “output” from language element “AbstractActivity” to “Alpha”.
• Add associations “input” and “output” from language element “Activity” to “WorkProduct”.

ActivitySpace is not in the position to create Alphas, it is merely importing them, acting like a containment, not a factory. Activities, on the other hand, are capable to influence the lifecycle of Alphas. Rearranging the Associations would distort this distinction. Also, the suggested removal of Action would remove the ability to specify the intended action on Aplhas and WorkProducts. The model as it stands is correct and shall not be changed.

Problem to be solved: It can be considered useful in practice to have not only completion criteria for activities, but also entry criteria. While the spec supports the former, it does not support the latter.

Proposed solution: Add optional entry criteria. Language element “CompletionCriterion” can be renamed to “Criterion” and be reused for that purpose. Possibly card updates in section 9.7.6.8 (pages 132-133) are necessary to make clear how entry criteria are shown on the cards.

The issue author raises a valid point. Change the preexisting CompletionCriterion into an abstract and neutral Criterion, which then can be specialized into an EntryCriterion and CompletionCriterion.

Problem to be solved:

It is currently not possible to attach resources, patterns or anything else to approaches or refer to approaches in patterns, as they are a plain string attribute of Activities instead of being a language element.

This makes it difficult to complement an approach with step-by-step instructions, scripts, tool mentors, examples or any other forms of guidance.

Proposed solution:

Replace attribute “approach” on langage element “Activity” by a new language element “Approach” (direct subclass of “LanguageElement”) with attribute “name” of type String. Add an association from “Activity” to “Approach” with multiplicity 1..* at the “Approach” end.

Agreed. Accept the proposed resolution.

There is definitely a useful state between conceived and coherent. Some people think bounded is too waterfall but accept the need for a state between Conceived and Coherent. We offer the following list for their consideration as possible, but not necessarily better, alternatives to “bounded”.
The candidate list of potential alternatives is (in no particular order):
Understood, Clarified, Themed, Scoped, Profiled, Envisioned
None of the new proposals are considered obviously superior to “bounded”

It has not been possible to reach a consensus on an alternative name. So the name will remain "Bounded".

Problem to be solved:
Acknowledgement missing in page 192 section B.1.1

Proposed solution:
Complete the following paragraph:
Hanna J. Oktaba and Miguel Ehécatl Morales Trujillo lead the work on the KUALI-BEH Kernel extension, which was based on the KUALI-BEH 1.1 revised submission guided by Hanna J. Oktaba

Adding:
... and Miguel Ehécatl Morales Trujillo, and on the KUALI-BEH 1.0 initial submission with participation of Magdalena Dávila Muñoz.

Agreed.

Problem to be solved:
The terms and definitions, presented in section 4, must be consistent with the ones used in the specification, particularly:

• Activity
• Competency
• Method
• Practice

Proposed solution:
Replace the definitions of those terms in section 4 by the definitions presented respectively in:
9.3.4.4 Activity
9.3.5.2 Competency
9.3.2.11 Method
9.3.2.14 Practice

Replace the first sentence of each of the four definitions in clause 4, with the first sentence of the corresponding description subclause in clause 9 for that term.

Problem to be solved: The Alpha class does not capture the essential qualities of an Alpha in a particular attribute. Thus section 9.7.4.7.3 should stop referring to, and displaying, essential qualities as though they were a referencable property of an Alpha. For example, they are shown on the sample Alpha card, Figure 53 in Section 9.7.4.7.3 Alpha Definition Card on page 118.

Proposed solution: The offending references are removed and the card shown is replaced by one that is conformant with the language.

The proposed solution is agreed to.

Problem to be solved:
Both words are correct, however the word "coherency" is more common and more used in the specification.

Proposed solution:
Replace word "coherence" by "coherency" in:

• page 204 Table 48, third row, first bullet
• page 206 Table 49, second row, third column, first paragraph
• page 220 subsection How Method Enactment drives the Work, fourth paragraph

(Note: The page numbers in the issue description are the sequential page numbers from the cover page of the document, not the page numbers as given in the page footers.)

It is not agreed that "coherency" is more common than "coherence". Indeed, online dictionary definitions seem to consistently define "coherency" as meaning "coherence", indicating that "coherence" is the primary term.

However, the document current uses both "coherence" and "coherency". For consistency, only one should be used, and that should be the primary term "coherence", not "coherency".

Problem to be solved:
The word "completeness" has a very specific meaning in mathematics and formal languages and needs a formal demonstration, while the word "sufficiency" is enough to describe this particular method property.

Proposed solution:
Replace word "completeness" by "sufficiency" in:

• page 83 section 9.3.2.11 (twice)
• page 200 section B.1.2.3 (twice) in: Description and States subsections
• page 201 Figure 113
• page 202 section B.1.2.3 Progressing the Method Authoring
• page 203 section B.1.2.3 (thrice) Progressing the Method Authoring
• page 203 Figure 116
• page 204 Table 48, third row, first bullet
• page 206 Table 49, second row, second column, first paragraph
• page 206 Table 49, second row, third column, first paragraph
• page 206 Progressing the Method Enactment
• page 220 subsection How Method Enactment drives the Work, fourth paragraph

A lot of words have specific meanings in "mathematics and formal language", but they also have clear meanings in less formal language. Further, in the context of the use in question, "sufficiency" does not have the same connotation as "completeness" for the usual reader. The definition of the "completeness" property currently is that a method "is complete if the achievement of all practice objectives fulfills entirely the method purpose and produces expected output". The alternative of saying that a method is "sufficient" does not seem to capture nearly as well as "complete" the idea of "fulfilling entirely the method purpose".

Problem to be solved:

The spec does not allow for checklists on a competency level or the creation of a competency level detail card.

Proposed solution:

Add an association “checkListItem” of type “Checkpoint [*]” to language element “Competency level”. For architectural clearness, move language element “Checkpoint” from package “AlphaAndWorkProduct” to package “Foundation”.
Add new section “9.7.6.8.4 Competency Level Detail Card” similar to section “9.7.4.7.5 Alpha State Detail Card”.

Agreed. Accept proposal.

Problem to be solved:

The spec supports state cards but not level of detail cards.

Proposed solution:

Add new section “9.7.5.8.2 Level of Detail Card” similar to section “9.7.4.7.5 Alpha State Detail Card”.

Problem to be solved:
The definition of a practice establishes that it has a purpose.
According to the Practice definition, it has an objective, not a purpose. The method is the one that has a purpose.

Proposed solution:
Replace the word "purpose" by "objective" in pages 20 (section 4) and 24 (section 7.3).

Agreed.

Problem to be solved:
The subclause reference of the Method definition is missing.

Proposed solution:
Since other terms have a reference to their respective subclauses, (See subclause 9.3.2.11.) should be added.

Agreed.

In clause 9.7.2 on page 112, Figure 36 shows the Graphical Syntax, or Diagram Interchange Metamodel, residing in Package DiagramInterchange. This package and its content is missing from the metamodel in the XMI file (as well as from the ancillary MagicDraw model).

Proposed Solution:
Add Package DiagramInterchange and its content as shown in Figure 36 to the model.

Update the model with elements shown in Figure 36 and replace Figure 36 with a version (from the model) that has the UML errors corrected.

Element Action in package ActivitySpaceAndActivity is specified to perform "create, read, update, delete". Therefore attribute "kind" should be of type Enumeration instead of type String.

Proposed Solution:
in ActivitySpaceAndActivity::Action,
change the type of attribute "kind" to ActionKind.

Define a new Enumeration ActionKind with enumeration literals "create, read, update, delete".

Agreed, accept the proposed change. The transition from a String to an Enumeration improves the precision and expressiveness of the model.

Problem to be solved: The first OCL expression on kernel starts with "self.elements" which is undefined on kernel.

Proposed solution: Replace "self.elements" by "self.referredElements->union(self.ownedElements)".

Agreed.

Problem to be solved: The OCL invariant on LanguageElement uses incorrect references to "owningAssociation" towards its end. This is incorrect, because owningAssociation will never point to e1 or e2, which are LanguageElements.

Moreover, it uses reference "a.member", which is undefined.

Proposed solution: Replace "p1.owningAssociation" by "p1.languageElement" and "p2.owningAssociation" by "p2.languageElement".

In addition, replace "a.member" by "a.memberEnd".

Agreed.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause: A.3.2.3 System Element

I strongly object to the term System Element used in A.3.2.3 System Element. I think it is a poorly defined term and seems to imply that a system is part of a software system. In systems engineering and in industry in practice, system is used to describe the larger component. INCOSE describes a system in the following way: "An interacting combination of elements viewed in relation to function (INCOSE)". It then goes on to say: "The elements, or parts, can include people, hardware, software, facilities, policies, and documents; that is, all things required to produce systems-level results. The results include system level qualities, properties, characteristics, functions, behavior and performance." The subsequent definition in this specification only includes software, hardware and date. I think this is shortsighted. I recommend that a new term should be found such as software system element.

(From an evaluation comment by Matthew Hause.)

Even though, in Essence, a "Software System" is defined to include "software, hardware and data". this is still not as expansive as the INCOSE definition of "system" and the elements it includes. Therefore, it is reasonable to make it clear in A.3.2.3 that the elements being considered are specifically "Software System Elements".

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause: 8.2.4.3 Testing

Destructive testing concepts should be added to the following statement. It is mentioned above as a competency and should be repeated here as well.
"The testing competency encapsulates the ability to conceive and execute tests to demonstrate that the system is fit for purpose, usable, meets one or more of its requirements and constitutes an appropriate solution to the stakeholders needs."
(From an evaluation comment by Matthew Hause.)

The earlier part of the subclause does not refer to "destructive testing", but to the ability to "think destructively".

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

The definition of "Alpha" is confusing to me. The following two definitions are contained in the spec and are radically different:

Section 4.1 Alpha An essential element of the software engineering endeavor that is relevant to an assessment of the progress and health of the endeavor. Alpha is an acronym for an Abstract-Level Progress Health Attribute.

Section 8.1.2 What is in the Kernel? Alphas – representations of the essential things to work with. The Alphas provide descriptions of the kind of things that a team will manage, produce, and use in the process of developing, maintaining and supporting software. They also act as the anchor for any additional sub-alphas and work products required by the software engineering practices.

Also, an Alpha is far more that simply an “attribute” as defined in the name. The term "Progress Health Attribute" implies that it is a simple metric. Clearly it is far more than that.

(From an evaluation comment by Matthew Hause.)

Agreed that this is confusing.

I would suggest that the best way to do that is to update the bullet on alphas in 8.1.2, to adding to the end of the first sentence something like " and, as such, are relevant to assessing the progress and health of a software endeavor."

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause: 7.4.2

"The Essence Language emphasizes intuitive and concrete graphical syntax over formal semantics. This does not mean that the semantics are not as important or necessary. However, the description should be provided in a language that can be easily understood by the vast developer community whose interests are to quickly understand and use the language, rather than caring about the beauty of the language design. Hence, Essence pays extreme attention to syntax."

Reading the specification it is obvious that much care and attention was put into the definition of the language and its semantics. Also, making the statement that how things are arranged (syntax) is more important that what the mean (semantics) is an inappropriate selling point for a language. Both are equally important.

(From an evaluation comment by Matthew Hause.)

The quoted statement, as currently written, accurately reflects the intent of the designers of the Essence language.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause: 5.1 Symbols

"There are no symbols defined in this specification"
I found this somewhat confusing, as the specification is full of symbols. Possibly I do not understand what is meant by this section.

(From an evaluation comment by Matthew Hause.)

Clause 5 is an optional clause anyway. Since there are no symbols, it seems to make the most sense to just remove subclause 5.1 and make the clause include only the abbreviation currently under 5.2.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

I would have preferred a formal definition of Practice, Alpha and other terms prior to a discussion of the different levels of conformance and rigor of these elements. It would make it easier to understand what is meant, especially since the word practice is previously in several different contexts. Possibly referencing section 4 at the beginning would help. It was also difficult to tell when the terms were referring to parts of the Essence language or just using the terms in normal English. Also, the fact that Alpha is in fact an acronym but is used throughout the specification as the proper noun is also confusing.
(From an evaluation comment by Matthew Hause.)

The fact that there are terms used in a specific technical sense in the Essence specification is unavoidable – it is a large part of the purpose of any specification document to make exactly such definitions, and understanding this is necessary to understand the specification. The extensive set of definitions provided in Clause 4 is intended to help with this for Essence. The fact that the term "Alpha" was originally coined as an acronym does not seem particularly important in understanding its technical definition as given in Clause 4 and later use in the main body of the specification.

Placing the definition of conformance in Clause 2 before the definitions in Clause 4 is as required by the OMG and ISO specification document template, and this organization is true of all current OMG specifications. Nevertheless, it would not hurt if, as suggested, a reference is made at the beginning of Clause 2 to the definitions in Clause 4.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

In figure 9, what is the consequence of a stakeholder not being involved? Can I add that to the state machine if for example the person assigned to answer questions for a key organization refuses to respond to those questions?
(From an evaluation comment by JD Baker.)

This issue is on what is now Figure 6, the "States of the Stakeholders" in 8.2.2.1. Some endeavors may never progress the Stakeholder alpha to the Involved state. The main consequence of this would be that you could not then get to the Agreed state and it is likely that the endeavor will not meet Stakeholder expectations.

This is not necessarily something that needs to be explained in the specification, though. Perhaps it is more a users guide issue.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

It is not clear what a customer area of concern is. In section 8.1.4 it appears that the customer area of concern includes stakeholders. In section 8.2 that statement is made that "Software engineering always involves at least on customer" which begs the question, what is a customer? Is it a set of stakeholders and opportunities? In that case it would be a customer are of concern. Is customer an alias for Customer Area of Concern?

(From an evaluation comment by JD Baker.)

Regardless of what a "customer" is, the area of concern is defined in 8.2.1 as "This area of concern contains everything to do with the actual use and exploitation of the software system to be produced." In the sentence "Software engineering always involved one customer", the colloquial sense of "customer" seems largely sufficient – i.e., there is always some intended consumer expected to actually "use and exploit" what the software engineering endeavor produces. However, perhaps a small clarification on what a customer is could be helpful.

Problem to be solved: The definition of five generic competency levels in section 8.1.6 (pages 16 to 19) is not supported in that detail by the definition of the language elements in section 9.3.5 (pages 86 to 88). Mismatches are in particular:
- Section 8.1.6 states “Each competency has five levels of achievement. These are standard across all the kernel competencies […]”, while section 9.3.5.2 allows zero to many levels and does not enforce any standard instantiations.

• Table 1 in section 8.1.6 summarizes the proposed competency levels in checklist style, while section 9.3.5.3 does only allow free text properties for that purpose.

Proposed solution: It is proposed to enhance section 9.3.5 as follows:

• Change section 8.1.6 to say “across all competencies”
• Add a clarifying note in section 9.3.5 that each tool claiming conformance to the standard must implement the kernel and therefore has to provide the standard instantiations for competencies as defined in table 1 in section 8.1.6. To enforce this partially via the language directly, change multiplicity on association “possibleLevel” in section 9.3.5.2 from * to 5.
• Add an association “checkListItem [*]” in section 9.3.5.3 that targets model element “Checkpoint”. For architectural clearness, move this element from package “AlphaAndWorkProduct” to “Foundation”.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Figure 132 – SPEM Taxonomy of Core Describable Elements is hard to read. Recommend that the diagram should be cleaned up.

(From an evaluation comment by Matthew Hause.)

This diagram is simply copied from the SPEM spec, for purposes of reference. The resolution is similar to what is in the SPEM spec.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause: 8.3.4.2 Development

I couldn’t help but notice the emphasis on coding and programming in this section, with very little mention of design and no mention at all on modeling. Is Essence not meant to support MDA or MBD at all? If so, it should be specifically mentioned.

(From an evaluation comment by Matthew Hause.)

Essence as such is completely agnostic to MDA or MDD. The Development competency covers design, and modeling code be a practice used by those with this competency.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause: 8.1.4 Alphas: The Things to Work With

The following sentence does not make sense: "For example, the team performs and plans work does not imply any specific order in that they perform and plan the work. "

(From an evaluation comment by Matthew Hause.)

Agreed that the sentence needs to be edited.

Problem to be solved:

The visualization is counter intuitive as the shape seems to point from bottom to top while the arrows run from top to bottom.

Proposed solution:

Invert the shape

Agreed.

Problem to be solved:

The cards may not have enough space to display the full descriptions of all checkpoint in all cases. The language only defines one property “description” for checkpoints and does not support abbreviated descriptions. On the other hand, it declares attribute “name” which is not used on the cards or anywhere else.

Proposed solution:

The attribute “name” in language element “Checkpoint” is renamed to “shortDescription” for an optional abbreviated version of the full description. An OCL constraint is added to ensure that "shortDescription" is not empty only if “description” is not empty.
The definition of a card body (page 119) is extended by: “If provided, the abbreviated description of a checkpoint is used as a default.”

Problem to be solved: In both places the Completion Criteria: Requirements: Sufficient is wrong

Proposed solution: In both referenced cases change Completion Criteria: Requirements Sufficient to Requirements: Acceptable

Agreed.

Problem to be solved: There is no checklist item in the Architecture Selected state that ensures the key technical risks are agreed to that are referred to in the name of the state.

Proposed solution: Need to add one more checklist item to Software System Alpha, Architecture Selected State. Proposed wording for new checklist item is: Key technical risks agreed to.

The proposed solution is agreed to.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

Subclause 8.1.2

The definition of Competencies is self-referential: "Competencies –representations of the key competencies required to do software engineering."

(From an evaluation comment by Matthew Hause.)

In Clause 4, "Competency" is defined as "a characteristic of a stakeholder or team member that reflects the ability to do work" and "a capability to do a certain job". The description of "Competencies" in 8.1.2 should follow this definition, which would be non-circular.

Specification: Essence – Kernel and Language for Software Engineering Methods (ad/2013-02-01)

In defining stakeholders in section 8.1.4, the spec states "All stakeholders must be involved throughout the software engineering endeavor to support the team and ensure that an acceptable software system is produced." Keeping all stakeholders involved throughout any project is an ideal but in my experience, never happens, at least not in large projects. Is this a key factor in Essence or just highly desirable from a method/process perspective?

(From an evaluation comment by JD Baker.)

Saying "must" here is awfully prescriptive and, as indicated in the submitted issue, not always realistic. It is also stronger than the more detailed discussion on the Stakeholder alpha later in 9.2.2.1 would seem to support.

File: Essence Metamodel XMI (ad/2012-11-02)

The XML namespace URI for xmi is incorrect. It should be “http://www.omg.org/spec/XMI/20110701”.

Agreed.