Action Language Notes

Here are some rough notes on the proposed anatomy of the * UML action language. All semantic elements can be represented as text, a la SMALL, or as graphical symbols a la Scrall. For visualization purposes, graphical symbols are less restrictive than text statements. Eventually we will establish a clear connection between these elements and the corresponding UML 2.0 action semantics. For now though, we need to build something that works well, with a minimum of components.

A parser will yield the elements described here so, at this level, we don't care about notation or syntax. We do want to distill the essence of an action block specification so that we can build a machine that can execute any action block. Rather than build a full blown action language all at once, it is probably best to define the core elements and implement a primitive machine. Then, with some experimentation, we can extend or rethink the language to evolve something truely useful.

Ss the goal of these notes is sort out the concepts required to support the specification of action components as well as the run-time execution of those components in an action language interpreter. Whereas SMALL and SCRALL focus on syntax primarily, we need to sort out the underlying mechanics and integration with the metamodel. The next step will be to start modeling those concepts. Incremental versions of the action language metamodel will be posted at sourceforge for anyone interested.

Here are some questions we hope to answer:

• What configurations of data are exchanged? How many do we need?
• Do we have just one flow that can transfer any kind of data?
• Or do we create a separate flow type for each configuration of data?
• What kind of processes do we have? What are the inputs and outputs?
• ?

We must also keep in mind that varying degrees of parallelism are possible depending on the target architecture. So we must be careful not to define structures that perform only in a sequential context.

Key

Let's use the following symbols:

[ a | b ] - a or b
{ c } - zero, one or many c's
1{ d } - at least one d
< e > - zero or one e - optional
( d, f ) - d and f are grouped together in the same flow
g - exactly one - g is required

Action Block

An action block is a bundle of action language associated with a state, class operation or service. Action language consists of nodes (processes and stores) connected by flows.

Flows

A flow transfers data into, out of, or between nodes. Each flow has a unique numbered identifier and may be optionally named. Duplicate names in the same action block are allowed.

Every flow specifies at least one source and destination. The types and quantities of sources and destinations depends on the flow type.

Specification: Unique_id < Descriptive_name >

Object Flow

An object flow transfers zero or many object references. This flow can be generated by either a selector or a creator process.

Specification: None
Source: [ object_pool | selector | creator ]
Destination: [ object_pool | transmitter | eraser | tester ]
Run-time content: { object_reference }
Uses:

Populate object pool from selector or creator
Supply references to an eraser for deletion
Transfer object pool to other object processors (to be defined later)
Provide input to a tester to evaluate cardinality

Attribute Flow

An attribute flow specifies the transfer of one or more attribute name:value pairs. All attributes in a flow must belong to the same class. Each value must have a data type that matches that of its attribute.

Specification: class_name, 1{ attribute_name }
Source: [ object_pool | transform ]
Destination: [ object_pool | transform | tester | creator ]
Run-time content: { < object_reference > 1{ attribute_name:value } }
Uses:

Read attribute values using object references in an object pool
Write attribute values of a specific object
Write attribute values the same for all objects referenced in an object pool
Supply data to a transform or tester
Supply attribute initialization values to a creator (Cannot carry an object reference)

When writing to specific objects during run-time, the following data is expected:

{ object_ref 1{ attr_name:value }

Each flowing object reference is accompanied by one or more name:value pairs. To obtain the object reference, the origin of the flow must trace back to an object pool store or creator process.

An attribute flow into an object pool containing only name:value pairs and no object reference will write all objects referenced in the pool to the same values.

{ attr_name:value }

We could use this example to set

The following data streams when reading attribute values from an object pool:

{ object_ref 1{ (attr_name:value) }

Here we see that each data item includes an object reference with at least one accompanying attribute name, value pair. The syntax above does not suggest any particular implementation. You don't have to send an attribute name with each value. But there must be some mechansim to ensure that the attribute of each value in an attribute flow can be determined during runtime.

Value Flow

A value flow transfers a single value not associated with any attribute. Each value conforms to a defined user or core data type. Consequently, a value may be arbitarily complex.

Specification: None
Source: [ value_store | transform | service ]
Destination: [ value_store | transform | tester | service | selector ]
Run-time content: value
Uses:

Transfer a temporary value in and out of a value store
Output a value from a class operation, transform or service
Transmit a value to a transmitter for inclusion in a signal
Transmit a comparison value to a selector criteria expression

Status Flow

A process executes when all of its required input data is available. This means that multiple processes could potentially execute in parallel. But we might want to ensure that certain processes wait, even though data is available, until that data is refined further. Or an upstream decision might preclude invocation of a certain process.

A status flow communicates a boolean condition to a downstream process to effect or negate execution. If a status flow is enabled during run-time, the condition is enabled. A process may not execute until all input control flows have been enabled and all input data is available.

Before considering the run-time content of a status flow, we need to understand how flows are managed in general. Every flow, regardless of type, starts off in an "inactive" state. This means that the flow has not yet been populated with any data. Once a flow is populated by whatever process or store provides its input, the flow enters the "activated" state. This means that data is now available. The content of an "active" flow may be "empty". (An object selection that returns no references, for example, would populate an object flow with zero object references). Once all data in a flow has been consumed, the flow enters a final "consumed" state.

When a status flow leaves its "inactive" state, it must be set to either true or false. Any process receiving the false input will not execute. True means that the process may execute. The name of the flow reflects its condition when set. Let's say the name of the flow is "Max Pressure Exceeded" and it feeds into a process named "Close valve". The process may execute only if the value of the flow is True.

Status might be enabled repeatedly for a series of object references. The example taken from figure 6.16.1 of the SMALL paper shows this situation. Crates are sorted into large and small sizes. A flow of object references in SMALL, each with an attached guard condition, flows to two separate attribute write processes.

Each status flow in this example flows object references as well as status.

Specification: None
Source: tester
Destination: any process, including another tester
Run-time content: <>

There are some cases where, instead of a status flow, a value flow with a boolean data type could be used to similar effect. But it would not be possible to preclude execution of a downstream process with this method.

Nodes

A node is either a Store or a Process. Some configuration of data can flow into a node, out of a node or in both directions.

Stores

A store is a node where data is organized for access by one or more flows. It is more accurate to say that a store makes data available than to say that anything is actually being stored. They are freshly initiated when an action block is invoked and destroyed when the action block completes. So the scope of a store, regardless of type, is the current action block.

Object Pool

An object pool stores a set of object references. The object references may be created by the output object flow of either a selector or an object creator. Every object pool has exactly one object flow input. During run-time, an object pool may receive zero, one or many object references. So it is possible to end up with an empty object pool after the object flow is activated.

Any number of attribute flows may enter an object pool. Each such flow specifies a write to the attributes of any objects referenced in the pool. Also, any number of attribute flows may stem from an object pool. Each such flow specifies the reading of attribute values from referenced objects.

Flows In: { attribute } { object }
Flows Out: { attribute } { object }

Each action block in an instance based state has automatic access to an object pool initialized to contain a reference to the current instance. This special object pool is always named "self". Not sure whether we need an object flow to fill this store or whether we make a special rule and say that the self store has no input.

Value Store

A value store holds a single value defined by a data type. Consequently the content of a value store can be arbitrarily complex. A value store is populated by at least one value flow. Any write to a value store replaces any prior value. A value flow that exits a value store reads the current value.

A parameter value in a state action block or a class operation has no input value flow. The value is automatically made available when the block is initialized.

Flows In: { value }
Flows Out: { value }

Processes

A process is a specialized activity that has exercises some influence on metamodel content. The inputs, outputs and activity depends on the process type.

Selector

A selector takes as input a class, a quantity [ 1 | * ], an optional criteria expression and an optional path expression. It locates the specified object or objects in the designated class and outputs an object flow populated with matching object references. If no matching objects are found, zero references are produced in the object flow.

Flows In: { status } { value }
Flows Out: object
Specification: ( class_name ) ( quantity ) < criteria > < path >

Creator

A creator makes a new instance of an object. A reference to the newly created object may be provided by an output object flow. An input attribute flow can be used to initialize any of the object's attributes.

Specification: class_name
Flows In: { status } < attribute >
Flows Out: < object >

Eraser

An eraser deletes any objects referenced in an incoming object flow.

Flows In: { status } object

Transmitter

A transmitter accepts an object flow and value flow for each parameter. A signal is generated to each object referenced in the flow.

Specification: event_specification { value_flow_name, parameter_name }
Flows In: { status } { object } { value }

Tester

A tester accepts status, value and/or attribute flows and produces one or more status flows.

Specification: Internal comparison expressions (NOT an action block)
Flows In: { status } { value } { attribute }
Flows Out: 1{ status }

Transform

A transform is a process that, as its name implies, transforms one or more input values to produce a one or more output values. Calculations may be performed, data manipulated and coverted, values tested and assignments made. A transform cannot directly access the class or state models. This means that objects (including "self") may not be selected, attributes read or written or signals generated. Local value stores may be allocated and accessed. A transform cannot produce a status flow - you need a tester process for that.

Specification: Internal transformation expressions (NOT an action_block)
Flows In: { status } { attribute } { value }
Flows Out: { attribute } { value }

Example:

Transform Compute_volume

Inputs( Attribute_flow:Dimensions )
Output( Attribute_flow:Volume )

Volume.Volume = Dimensions.Length * Dimensions.Width * Dimensions.Depth
Volume.Object = Dimensions.Object

In the example above, a transform named Compute_volume is defined. It's single input is declared as an attribute flow. The local name Dimensions is applied. An output attribute flow is declared and named Volume. The L, W, D components of the input flow were defined outside the transform, but can be referenced internally. The single component V of the output flow is dynamically declared internally. It can then be referenced downstream. The object reference received in the Dimensions attribute flow is passed along to the Volume attribute flow.

Service

A service is a blackbox process provided by either an external service domain or in the caller's domain. No model references such as objects or attributes may be exchanged with the caller. Only status and values may be communicated.

Flows In: { status } { value }
Flows Out: { status } { value }
Specification: action_block

Operation

An operation is an activity defined on a class independent of any state chart. Operations are object-based so that a separate context executes for each object. All selected object references remain available to any invoked operation. This makes sense since the object references in an object pool are valid until the state action completes and the operation must return before that. An operation may produce a single value output. (When we design the exception throw-catch mechansim there will be an additional way for an operation to pass data out).

Flows In: { status } { value } { object } { attribute }
Flows Out: { status } { value } { object } { attribute }
Specification: action_block

Conclusion

Again, these are just rough notes in preparation for the modeling activity. As usual, the process of modeling will ferret out inconsistency, ambiguity, self contradiction and gaping holes in logic. All of which are surely present! Nonetheless, I thought it might be helpful to expose the current thinking. Feel free to comment if you see any trouble spots or feel I'm heading down the wrong road or have any theory to contribute.

- Leon

Still analyzing...

I had hoped to post my premilinary action language metamodel notes by now, but the quicksand keeps shifting! Still resolving inconsistencies and trying to boil down the necessary concepts. Getting close, though.

Small updates

I haven't posted for a few days so I thought I would just make a quick update. I am currently putting together some notes so that I can start building the action language metamodel. Naturally, you can't model what you don't understand and until I get a grip on this thing there's just not much to model. The good news, though, is that I'm making a lot of progress which I will post shortly.

I have been reading, re-reading, re-re-re-reading every little detail in the SMALL language paper and putting together some notes. As I progress I am finding a number of issues left, intentionally or not, as "exercises for the student".

Primitive processes such as "attribute write", "object create", etc. are pretty clear. But then you see something like "Compute Volume" (figure 6.16.1 on page 23) with an input data flow labeled "N" and an output labeled "Volume". Okay...

Clearly, somewhere inside that process, is the simple v = l * w * d expression. But we see that there will be one (l, w, d) for each Crate object - thus the N on the input. There's no N on the output! But the example doesn't work if we output only one v, so N must have been intended. Now we could say that the action language interpreter automatically binds the output to an object reference since it sees that the input flow carries an object reference. But what happens if we have an process that inputs multiple <...>N flows - each from a different class? Do we merge the output somehow? Or what if we really did intend to output a single value unassociated with any object reference? My point is that there is a lot of room for ambiguity here and the SMALL paper does not spell any of this out.

Now that's just one example. This paper is replete with missing pieces - all due respect to the authors! So I will make my best effort to fill them in, present some notes, and see what you all think.

Action Language Metamodel

I've been pondering the action language metamodel the last couple of days. For reseach, I've printed out the in-depth SMALL paper from back in 97 and xeroxed the two page summary from Executable UML as well as my SCRALL notes.

Goals of this action language are:

1. Emphasize parallel processing and data flows.
2. Ensure that the language is as easy to read as the class and state models.
3. Support a full range of data types and domain bridge mechanisms.
   

I'm really impressed with the quality of the SMALL paper. I will probably just implement it straight out. But how?

I certainly need to model action behaviors (link, create, write attribute, etc.) as primitive operations in the class and state metamodel subystems. For now I will just imagine these as domain functions with appropriate signatures. Example:

::Unlink(object_id1, object_id2, rnum)

Clearly we need a lex-yacc generated parser, which means I need to write up a full grammar. I'm thinking that the intermediate representation produced by the parser will populate an action language metamodel - yet to be built.

So when action language is entered, it will be parsed and the action language metamodel will be populated. When a block of action language is triggered at run-time, the action language metamodel will invoke the primitive operations with appropriate parameter values.

The lex-parser and primitive ops are the easy parts, so I will save those for last. I hope to post progress here in the coming days.

It's a nice day for a walk so I'm heading to the local cafe to ponder further.

First Data Type Class Model

The good news is that I've sorted out some terminology since the last post (as promised!). The bad news is that I've clearly got some more studying to do before nailing this data type thing down the way I want. I suppose, though, it's good news that I've figured that much out. After my conversation with Andrew, I'm concerned that this subsystem could turn into a serious project of its own. So we will probably limp along with a miminal data typing system whilst the rest of the action language gets sorted out.

For now, I propose the following model be inserted into the datatype subsystem of the * UML metamodel.


If you don't have an image resizer plugin (firefox) the diagrams might be hard to read. I'm new to the blog - writely technology and haven't figured out the easiest way to insert readable model fragments. I will post the full pdf and BridgePoint files on the sourceforge site, though. Probably in a separate file package.

Here I just want to excerpt and explain what I have so far. This is only a model of data type structure, not data type operations. The model is best explained in two pieces: core and user data types. Then I will construct the Position data type to show how the model works.

Core Data Types

First let's examine the left side - the core data types.

Nothing mysterious here. I have superclassed the Real and Integer Numeric types because they each may specify units.

Note that the max/min values on Integer and Real are kept in the subclasses since the types are not the same. The real type also specifies a precision as a positive integer representing decimal places to the right.

The Enumerated data type consists of a set of Symbols, each of which has a Value.

Finally, String has a positive integer maximum length. Now let's flip over to the right side.

User Data Types

A Vector is simply a collection of one or more values (as determined by Vector.Scale) all specified by the same Datatype. Since the Datatype can be either User or Core we can set up a hierarchical structure. Since Vector.Scale can be 1 or more we can build an N-dimensional matrix at any level of the hierarchy. Let's review the examples mentioned in my earlier post.

First we'll work from the bottom up to define the Position data type. Since minutes and seconds are the same in both latitude and longitude, let's define those.

Minutes

Our specification for Minutes is:
Name: Minutes
Core Type: Integer
Min_value: 0
Max_value: 59
Units: minutes

So this gives us an instance each of: Integer, Numeric, Core Data Type and Data Type

Seconds

Using a more compact syntax, we specify:
Seconds: real[0..60] prec 2 units seconds

This gives us an instance each of: Real, Numeric, Core Data Type and Data Type

Degrees Latitude

Lat_Degrees: int[-90..90] units degrees

Degrees Longitude

Long_Degrees: int[-180..180] units degrees

Unlike thte earlier post I am using negative degrees to avoid defining an enumerated type for [N | S] and [W | E].
No particular reason - just lazy. Also the max Minutes is 59 while max Seconds is 60 (59.999...). Seems like it is better to say "60". Otherwise you end up redundantly specifying the precision with a sequence of trailing nine's.

Latitude and Longitude

Now we start creating User Data Types as we work our way up the hierarchy.

Name: Latitude
Vector: name Degrees, scale 1, value_datatype Lat_Degrees
Vector: name Minutes, scale 1, value_datatype Minutes
Vector: name Seconds, scale 1, value_datatype Seconds

Name: Longitude
Vector: name Degrees, scale 1, value_datatype Long_Degrees
Vector: name Minutes, scale 1, value_datatype Minutes
Vector: name Seconds, scale 1, value_datatype Seconds

Position

Name: Position

Vector: name Latitude, scale 1, value_datatype Latitude

Vector: name Longitude, scale 1, value_datatype Longitude

Matrix example

Now if, for some reason, we wanted a 5x10 matrix of Position data we could do this:

Name: Position_Column

Vector: name Position, scale 10, value_datatype Position

Name: Position_Matrix

Vector: name Position_Rows, scale 5, value_datatype Position_Column

Moving on

Okay, that's it for now. I might shelve the data type subsystem for a few days while I get some other tasks under control. Namely: Publish up a project roadmap, resume work on the model build script parser, sketch out key parts of the action language subsystem.

Stay tuned!

- Leon

Data Typing Notes - Part 2

Operations on Core Data Types

Before moving on to the interesting part - application of operations on composite data types - let's take stock of what we already know about how operations are applied to core data types in programming languages.

A built in operation may be applied to core data type values using one of three common formats:

• Binary operator
• Unary operator
• Assignment operator
  
• Function call
  

We generally see the binary operator format in math expressions like this:

x = 2 + 8

It is rarely done, but we could rephrase the above statement using the function call format:

x = add( 2, 8 )

Generically, these formats are:

x = a OP_SYMBOL b
x = OP_NAME( a, b )

where a, b are the values and x is the result.

Finally, we have a couple examples of the unary operator format:

++c
-d

Again, we can restate both examples in the function call format:

c = increment( c )
x = negate( d )

In addition to the obvious fact that each function has only one input value, we observe that the first example changes the input value while the second does not.

The assignment format modifies the actual value

x *= 2
x OP_ASSIGN a
x = OP_ASSIGN_NAME( x, a )

Okay, let's set those observations aside and take a look at a composite data type and see what kind of trouble we can get ourselves into.

Operations on Composite Data Types

Whenever we create a new composite data type, no matter how simple, we need to review all of the standard operators and consider what functions make sense on the new type. Consider an application where you need to count how many times something occurs. Maybe we have an automotive application and we need to count how many lubrication cycles occur within a given time frame. This sounds like a job for an integer, but not quite. What we need is a counter. It can't ever be negative. Operations-wise we want to increment it, decrement it and reset it to zero. Let's define it.

We use a hierarchy format:

Layer : Counter

count : int[0..max]

So we have a data type structure named "Counter". To define our operations we use the OOP mechanism of inheriting the operations already defined in the integer base type with some overriding features. Specifically we want to use the ++ and -- unary operators, but we need to make sure that -- never takes us past 0. Additionally, we need a reset function. Rather than use a hard to read/remember operator we can stick to a function name.

Counter::

op(--) :
if --count < 0, count = 0
op(*) : not defined
op (-) : not defined
op (+) : not defined
reset () : count = 0

Please don't pick apart the syntax above. I'm not trying to nail that down yet! We're really just focusing on the required data. The example above shows that unless otherwise specified, all int operators and functions apply. We show that -- works differently. Some action language will be written in a function to be attached to the -- operator to override normal -- behavior. We turn off the *, - and + operators. We could just leave them on but let's say that we want to tightly define counter behavior. It's a counter and nothing else! A new function is defined for reset. All of this is pretty standard OO inherit and override behavior except that Counter is not an object or a class - it's just a data type in * UML. The distinction will be important and this IS different than OO programming languages.

The binary operation takes two inputs and returns a result without changing either input, so we need to specify a little more:

Compass_Heading::

op(+) :

sum = a + b
if sum == 0:

return 0

else:

return sum % 360

op(-) :

# insert action language here

op(*) : not defined

In the above psuedo-syntax we use the letters a and b to represent the symbols left and right of the operator. The result to be assigned to the LHS of the expression is "returned". By the way, I am imagining a GUI that displays all operators defined for the core type as check boxes that can easily be flipped to inherit_none or inherit_all. Then the modeler can click the boxes that are exceptions to the rule (inherit or don't inherit). To preclude weird analysis bugs, it is just as important to undefine operators as it is to extend them!

I just had a conversation with my colleague and good friend Andrew Mangogna and realize that I'm starting to confuse my terms regarding hierarchies, layers, matricies, etc. I am going to get this all sorted out by the next post.

Data Typing Notes

The term "data type" describes both a structure and a set of operations. Consider the integer data type. Transforming operations such as add, subtract and multiply yield a new integer value. Comparison operations like greater_than, less_than and is_equal_to yield a boolean result.

Analysts rely on two kinds of data types during model development. Core data types and composite data types. Core data types are built in to the modeling language. The analyst can create a composite data type using one or more core data types as building blocks.

What core data types should * UML support? Is the list of core data types bounded by some principle or can you just keep adding them? How are composite data types actually built? How much structure is allowed in a composite data type - is there such a thing as too much? How are composite data types handled by model compilers? How are they handled by an interpreter? These are just a few of the questions that must be answered to construct a useful * UML Data Type metamodel subsystem.

Core Data Types

Since a core data type is built into the modeling language it must be supported by any model compiler. Both the data type structure and all operations should be fully implemented. Keep in mind that a set of models might be translated to Ruby or Java on one day and then directly to Z80 assembler on another. To support all model compilers, we are safer to keep the range of core data types as limited as possible.

On the other hand, model developers like to have a wide variety of data types. In fact, real world applications rarely call for a core data type as seemingly ubiquitous as real. Consider an air traffic application. The compass heading data type is not real. It is in fact [0..360] degrees with a precision of, say, .01. The math operations aren't the same either for compass heading. 180.15 + 270.00 = 90.15. Similarly we need distinct data types for altitude, speed and so forth. True, each of these user data types can be tightly or loosely based on the real core data type, but each is different.

So we have a balancing act. To model application requirements independent of any implementation, we want to support lots of user (application level) data types. To support a wide range of model compilers and target languages, we must limit the set of core data types. And right there we have the answer to our problem.

Any user data type must be derivable from core data type building blocks. The model compilers will be concerned only with the core data types. But the modeler will have everything he or she needs to specify whatever application level data types the project demands. Before moving on to how the user builds composite data types, lets first consider the core set.

Consider these proposed core data types:

• boolean
• integer
• real
• enumerated
• string

The first two types, boolean and integer, are supported in every programming language down to assembly level (as far as I know). So they are safe bet. At the level of C and upward, the real (float) data type is supported. Some embedded processors lack floating point capability. But floating point math is quite common in real world applications. In the worst case, a model compiler will have to convert all read core types into integer types during compilation.

Now we move on to enumerated. This is easily implemented using integers so we easily demand this capability of any model compiler.

What about the string data type? C doesn't support strings directly - why should * UML? The data structure is easy enough to accommodate, but there are a number of string operations that must be supported. But strings are extremely common in real world applications. It's hard to imagine avoiding them. That said, certain embedded targets have resource limits that preclude the liberal use of string data types. A model compiler might need to convert string values into a concatenated or compressed integer form to acheive any necessary economy. Since I'm an analyst I have to say "enough for me, let's shift the burden onto the model compiler!" We need strings!

But that's it - we're drawing the line at string. You want to build a model compiler? These are the only types you need to handle. I now assert that the modeler has enough building blocks to construct any required user data type.

Composite Data Types

To build a composite data type we create a Type Specification using one or both of the following layout patterns:

• Hierarchy
• Matrix

Hierarchy Layout

Let's say that we want to create the application level Position data type. Hang on - let me bring up Google Earth so I can make sure I get this right... Okay, the Golden Gate Bridge is here:

37 48' 59.41" N
122 28' 54.05" W

We create a Layer called "Latitude" with four data Cells. The cells are ordered sequentially as:

Layer : Latitude

degrees: int[0..90]
minutes: int[0..60]
seconds: real[0..60] prec 2
pole: enum( N | S )

Each cell refers to a core data type. Each core data type has its own qualifying characteristics.

Similarly we create another Layer called "Longitude":

Layer: Longitude

degrees: int[0..180]
minutes: int[0..60]
seconds: real[0..60] prec 2
meridian: enum( W | E )

It is clear we could have created a lower layer to contain minutes and seconds since they are identical, but let's not worry about that now. We create a top layer for Position and we have a complete Type Specification.

Layer: Position

Latitude
Longitude

It seems that composite data types don't require qualifying characteristics. Is this true? Hmmm.

Matrix Layout

We can also arrange core data type cells in a matrix pattern. To specify a 10x10x2 3D matrix of integers between 1 and 100 where any value is at (x, y, z) we could do this:

Matrix: int[1..100]

x: 10
y: 10
z: 2

Ah, but what if we wanted to create a 1x5 matrix of Position values for some reason? Then we combine both the Matrix and Hierarchy layout patterns.

Matrix: Position

p: 5

Composite Data Structure Summary

The Hierarchy and Matrix layout mechanisms make it possible to construct composite data types. But it's not enough to define a data type. We need to somehow define the operations specific to that data type.

But I need to get a sandwich right now. So stay tuned for the next post where we will ponder transformation and comparision operators on composite data types and such things.

- Leon

Getting Started

Welcome to my * UML blog.




* UML is a set of free (as in beer) UML metamodels and a supporting editor. * UML is a language for building models with the following characteristics:

• Essential application analysis/requirements are kept separate from the implementation mechanics. (I avoided the word "details" here because it is a myth that there are more details in implementation than there are in the application rules. The details just happen to be different).

• All models are fully executable. This means that if you build a * UML model you can say "run" and it will execute just like any program code. Even though the model contains nothing but UML level concepts like classes, attributes, states, and action semantics, the models run without the need for any inserted C, Java, C++, Python or any other alien program language code. Now the models might be running on top of a virtual machine built in one of those languages, but there is no code in the models themselves. This is important because we need to keep the models separate from the implementation.

• All models concepts are expressed in a maximally parallel form. In most programming languages instructions execute in sequence unless parallelism is made explicit with, say, threads or processes/tasks. In * UML it is just the opposite. Everything runs in parallel (on an imaginary massively parallel architecture) and sequence only occurs where explicitly specified. A key benefit of a language like this is that once you build a set of models for an Air Traffic Control Application, for example, you can compile the models to platforms with a variety of parallelism implementation. The same models could be compiled onto a single processor architecture or onto a distributed processor architecture. Same models, different code generation. After all, the application rules don't change - why should the models that formalize those rules change? Only the implementation goo should vary.

• The modeling language itself should be have as few symbols/rules as possible. The less the better. This is because we want the language mechanics to fade into the background with the rules of the application brought to the forefront. The goal of a * UML model is to express the application rules as clearly as possible, in as much detail as possible, without specifying an implementation.

* UML is inspired by these books if you want to know more:

Executable UML
ISBN: 0201748045
Model Driven Architecture with Executable UML
ISBN: 0521537711
MDA Explained
ISBN: 0201788918

The purpose of this blog is to post technical notes and developments as I construct the * UML editor and accompanying metamodels.

By the way, you can download the current release at SourceForge:

http://sourceforge.net/projects/starruml/

Finally, my company's web page is here:
http://modelint.com

That's all for now.

- Leon