UML Diagramming Guide

A Supplementary Guide to
Pacestar UML Diagrammer 5.0
 Pacestar UML Diagrammer
Version 5.0 for Windows

UML Diagramming Guide

IMPORANT NOTICE:
Copyright © 2005 Pacestar Software. All rights reserved worldwide.
Information in this document is subject to change without notice. No part of this document may be
reproduced in any form or by any means - graphic, electronic, or mechanical - including photocopying,
recording, taping, or storage in any information retrieval system, for any purpose other than to aid a licensed
user directly in the usage of the software, or when authorized in writing by Pacestar Software.
Pacestar Software retains all ownership rights to this computer program and its documentation. The source
code is a confidential trade secret of Pacestar Software. You may not attempt to decipher or decompile the
program or develop source code for it, or knowingly allow others to do so. Making copies of the program
(except for archival purposes or as an essential step in the use of the program) is prohibited. The program and
its documentation may not be sublicensed and may not be transferred without the prior written consent of
Pacestar Software.
Pacestar Software PROVIDES THIS PUBLICATION “AS IS” WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED. Pacestar Software may revise this publication from time to time without notice.
Every attempt has been made to assure that this manual provides the most current and accurate information
possible.

ACKNOWLEDGEMENTS:
All copyrights and trademarks mentioned herein belong to their respective owners. Pacestar Software is a
trademark of Pacestar Software. Windows is a registered trademark of Microsoft Corporation. Microsoft
Word is a registered trademark of Microsoft Corporation.
 Table of Contents 1

TABLE OF CONTENTS
About This Guide ............................................................................................................ 9
Symbols and Conventions.........................................................................................................9
General Techniques ..................................................................................................... 11
General Style Usage Tables ...................................................................................................11
Keywords.................................................................................................................................11
Attachable Nodes ....................................................................................................................12
Attaching Attachable Nodes to Base Nodes .....................................................................13
Repositioning Attachable Nodes.......................................................................................13
Detaching Attachable Nodes from Base Nodes................................................................14
Comments ...............................................................................................................................15
Containers ...............................................................................................................................16
Frames ....................................................................................................................................17
Path Labels .............................................................................................................................18
In-line vs. Lateral Path Labels...........................................................................................18
Flow Labels .............................................................................................................................19
Off Center Path Connections ..................................................................................................19
Path Trees...............................................................................................................................20
Creating Path Trees..........................................................................................................21
Manipulating Path Trees ...................................................................................................21
Node Symbols Containing Internal Icons ................................................................................23
Extensions and Nonstandard Symbols ...................................................................................23
Miscellaneous Drawing ...........................................................................................................24
Activity Diagrams ......................................................................................................... 25
Sample Activity Diagram .........................................................................................................25
Activity Diagram Style Usage Tables ......................................................................................26
Forks and Joins .......................................................................................................................28
Object Flows and Object Flow States .....................................................................................28
Conditional Branches and Decisions.......................................................................................29
Pins .........................................................................................................................................29
Listbox Pins .............................................................................................................................31
Connectors ..............................................................................................................................32
Exception Parameters .............................................................................................................32
Interruptible Activity Regions...................................................................................................33
Class/Object Diagrams ................................................................................................ 35
Sample Class Diagram............................................................................................................35
Class/Object Diagram Style Usage Tables .............................................................................36
Associations ............................................................................................................................40
Association Navigability ....................................................................................................40

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 1 of 4
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide TOC.fm
2 Table Of Contents

Association End Labels .................................................................................................... 41

Association Multiplicities ................................................................................................... 41
Association Names........................................................................................................... 42
Association Direction Indicators ....................................................................................... 42
Association Constraints .................................................................................................... 43
Association Classes ......................................................................................................... 43
N’ary Associations ............................................................................................................ 43
Qualifiers ................................................................................................................................. 44
Adding Qualifiers to Classes ............................................................................................ 44
Detaching Qualifiers from Classes ................................................................................... 45
Templates ............................................................................................................................... 45
Communication Diagrams ...........................................................................................47
Sample Communication Diagram ........................................................................................... 47
Message Style Usage Tables ................................................................................................. 48
Messages................................................................................................................................ 49
Creating Messages (Point-to-Point) ................................................................................. 49
Creating Messages (Stamping) ........................................................................................ 50
Labeling Messages........................................................................................................... 51
Looping Messages .................................................................................................................. 51
Component/Deployment Diagrams .............................................................................53
Sample Component Diagram.................................................................................................. 53
Sample Deployment Diagram ................................................................................................. 54
Component/Deployment Diagram Style Usage Tables........................................................... 55
Interfaces ................................................................................................................................ 56
Labeling Interfaces ........................................................................................................... 58
Moving Interfaces ............................................................................................................. 58
Assembly Connectors ............................................................................................................. 58
Ports........................................................................................................................................ 59
Creating Ports................................................................................................................... 60
Labeling Ports................................................................................................................... 60
Attaching Interfaces to Ports ............................................................................................ 60
Repositioning Ports .......................................................................................................... 61
Composite Structure Diagrams ...................................................................................63
Sample Composite Structure Diagram.................................................................................... 63
Interaction Overview Diagrams ...................................................................................65
Sample Interaction Overview Diagram.................................................................................... 65
Package Diagrams ........................................................................................................67
Sample Package Diagram ...................................................................................................... 67
Package Diagram Style Usage Tables ................................................................................... 68
Containments .......................................................................................................................... 69

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 2 of 4
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide TOC.fm
 Table of Contents 3

Sequence Diagrams ..................................................................................................... 71

Sample Sequence Diagram ....................................................................................................71
Sequence Diagram Style Usage Tables .................................................................................72
Lifelines ...................................................................................................................................73
Activations ...............................................................................................................................74
Terminate Nodes.....................................................................................................................76
Messages ................................................................................................................................76
Recursion and Subactivations.................................................................................................79
Guard Conditions on Activations and Lifelines........................................................................80
States on Activations and Lifelines..........................................................................................80
State Diagrams ............................................................................................................. 81
Sample State Diagram ............................................................................................................81
State Diagram Style Usage Tables .........................................................................................82
Superstates .............................................................................................................................84
Concurrency Boundaries.........................................................................................................84
Entry and Exit Points ...............................................................................................................85
Labeling Entry Points and Exit Points ...............................................................................85
Alternate Notation for Entry and Exit Points......................................................................86
Self-transitions.........................................................................................................................86
State Name Tabs ....................................................................................................................87
Use Case Diagrams ..................................................................................................... 89
Sample Use Case Diagram.....................................................................................................89
Use Case Diagram Style Usage Tables..................................................................................90
Relationships...........................................................................................................................91
Unsupported ................................................................................................................. 93
Timing Diagrams .....................................................................................................................93
Parameter Sets .......................................................................................................................93
Explicit Directions on Input and Output Pins ...........................................................................93
General Ordering Path ............................................................................................................93
Use Case Stereotype Symbol .................................................................................................94
Information Class Stereotype Symbol .....................................................................................94
Custom Stereotype Symbols...................................................................................................94
References .................................................................................................................... 95

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 3 of 4
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide TOC.fm
4 Table Of Contents

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 4 of 4
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide TOC.fm
 9

About This Guide

This UML Diagramming Guide is a companion to the Pacestar UML Diagrammer software.
Within this guide you will find techniques for developing and maintaining Unified Modeling
Language diagrams using the features, templates, and symbols contained in the software. It will
assume a basic understanding of the product and its diagramming features and terminology
documented in Pacestar UML Diagrammer User Guide. Both this guide and the main user guide
are duplicated in limited detail in online help modules accessible from within the software.
The version of Pacestar UML Diagrammer that accompanies this guide is based on UML 2.0 as
defined by OMG and published in the official specification, and taking into account some of the
more popular industry references. The notation we use for cross referencing the contributing
reference publications is noted below in Symbols and Conventions. Where conflicts,
discrepancies, or voids are apparent, we do our best to combine sources and weigh popular
acceptance and usefulness in determining how to support the notation and terminology. The
dynamic nature of UML and the abundance of tools and publications trying independently to pin
it down, results in constantly changing requirements. Consider this guide to be advisory and
certainly not authoritative in regard to the language notation. If any of the notation described
becomes obsolete or falls in disfavor with the industry, you can use the more generic capabilities
of the software to adapt.
How can I get the tool to something? Refer to the main User Guide
How can I diagram specific UML notation? Refer to this user guide
What is proper UML notation? Refer to OMG specification or leading references

Please note that neither our developers nor our technical writers claim to be experts of UML
usage. The descriptions, examples, and guidelines we provide are meant to be informative with
regard to the software, not necessarily representative of good (or even proper) UML usage.

Symbols and Conventions

z Names of keys on the keyboard are shown in small capital letters such as CTRL, ESC, and
ENTER. Key combinations are shown as CTRL-C, ALT-F, and so on. Remember that keys may
not be labeled exactly the same on your keyboard.
z Important terminology is italicized. Most terms are defined by UML, but some are our own
inventions. Many are also terms that have been adopted by the software and are defined in the
main user guide.
z Reference publications are listed in the appendix. When they are cross referenced in the text to
explain standards, ambiguities, or conflicts they are denoted by a bracketed symbol and
optional page number such as [SPEC231] (UML official specification page 231). The symbols
are included in the References chapter.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 9 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
10

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 10 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
1 GENERAL TECHNIQUES
This section describes techniques that are useful working with any type of UML diagram.

General Style Usage Tables

GENERAL PATH STYLE USAGE (COMMON TO ALL UML DIAGRAMS)

Style UML
Appearance Description
Name(s) Construct(s)

Anchor anchor Connection to a Note for comments and constraint nota-

tion. The anchor point shown here at the end is visible only
when not terminating on a path or node.

GENERAL NODE STYLE USAGE (COMMON TO ALL UML DIAGRAMS)

Style UML
Appearance Description
Name(s) Construct(s)

Note note Notes contain comments or additional information

concerning a diagram or any element of a diagram. Notes
can appear alone of be connected by an Anchor path to a
diagram element or an area of a diagram.

Frame frame A container for identifying related diagram elements.

Frames have specific uses within diagrams such as com-
bined fragments in sequence diagrams but they are avail-
able in all UML diagrams. One common use is to bound
any diagram or diagram fragment to identify the type of
diagram, such as class, component, communication,
sequence and so on.

Keywords

UML keywords are enclosed in guillemets (double angle brackets). These characters are not available
on most keyboards. To add them to text, type two consecutive single angle brackets “<<“or “>>”. The
software will automatically combine these into a single guillemet. To circumvent this behavior, type a
space between the two, then after typing the second one, go back and delete the space - just so long as
the two identical angle brackets are not typed consecutively.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 11 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
12 General Techniques

Attachable Nodes
UML notation contains a wide variety of nodes that attach to other nodes in different ways. For
example, qualifiers attach to the outer edges of class nodes, ports attach to components, input and
output pins attach to actions, and so on.

Pacestar UML Diagrammer allows you to attach certain nodes to others so that they can be
manipulated together. In most cases, this relationship involves a base node and an attachable node
that will attach to the base node. A port, for instance is an example of an attachable node that can
be dragged to a component base node where it will snap into place and become attached to the
component. The inverse is not true. A component cannot be dragged to attach it to a port. Most
attachable node relationships have this sort of one-directional relationship with their base node. A
notable exception are lifelines and activations which can be attached to objects bidirectionally.
That is to say that objects can also be attached to activations or lifelines. The process of attaching

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 12 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 General Techniques 13

nodes to one another is designed to be intuitive for the most part and to reflect the way that you
would construct diagrams on paper.

Attaching Attachable Nodes to Base Nodes

% To add an attachable node to base node

We’ll assume the base node is already on your diagram. In the case of a port that is to be attached
to a component, the component is the base node, and the port is the attachable node.
1. Select the attachable node symbol from the style bar (a port for example). Be careful to select
the proper attachable node for the base. It’s easy to confuse a port with a pin, but a port will
only attach to a component and a pin will only attach to an action.
2. Move the cursor near the location of the base node where the attachable node will attach. The
particular position depends on the relationship between the attachable node and the base node.
In the case of a port and a component, the port attaches to any point overlapping the edge of the
component, though it can also attach just inside the edge in the less common case of a private
port. The software knows which nodes attach to which, and where they can attach.
3. When the cursor is located at a valid location where the attachable node is permitted to attach
to the base node, the attachable node will snap into place. If for some reason you prefer for the
node NOT to attach to the base node, simply hold down the CTRL key and it will not snap into
place (nor will it become attached if you create it).
4. Click to create the attachable node on the base node. The attachable node (the port for
example) will drop into place and attach to the base node (the component).
Once an node is attached to a base node, the node will remain attached to the base node when it is
moved, copied, and duplicated. It will also be deleted when the base node is deleted. If you prefer
to delete the base node not the attached nodes, simply detach them first.

NOTE: Some attachable nodes, notably the listbox pins and complex ports, have separate
vertical and horizontal configurations (different styles). These can only be attached to
the appropriate side in the way you might expect. The horizontal version of the symbol
can only attach to the top and bottom edges of the base node, and the vertical symbol
can only attach to the left and right sides.

Repositioning Attachable Nodes

After a node is attached to a base node, you can select it individually and drag it to reposition it to
a different position on the base node. You can also move it away from the base node to detach it.
Note that one exception is made for activations and lifelines and their attached objects. Because
these relationships are bidirectional, moving either will move the other rather than repositioning.
However, these nodes can only be attached in one way (top to bottom) so there is no need to
reposition either.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 13 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
14 General Techniques

For example, to reposition a port that is attached to a component, simply select only the port and
drag it until it snaps to an alternate location on the edge of the component. You can accomplish
this by dragging the port when nothing is selected as well. If you select the port, but do not select
the component, AND you select any other objects, the port will not move because the component
to which it is attached is not being moved. While this is usually expected, you might be surprised
to find that it also applies to other ports attached to the same component. If you attempt to drag
more than one attached node at a time (with the intent of repositioning both) neither will move.
Simply reposition one at a time instead.

NOTE: When duplicating (via CTRL-drag) an attached node, be sure to release the CTRL key
before dropping the copy. Otherwise, the copy will not snap into place. The CTRL
serves dual purposes, to create a duplicate, and to invoke freeform mode where the
object being manipulated will ignore everything else in the diagram (will not attach
when dropped for example).

Detaching Attachable Nodes from Base Nodes

A node that is attached to a base node can be detached so that it becomes independent of the base
node. Once detached, it will no longer move along with the base node, be copied with it, or be
automatically deleted when the base node is deleted. There are two ways to detach most
attachable nodes from their base nodes.

% To detach an attachable node from a base node (method 1)

As described in the previous section, the easiest way to detach an attachable node from a base
node is simply to select it individually and drag it away from the base node. Once it is out of
range of the base node, dropping it will sever the relationship with the base. Be sure to drag the
attached node when either (a) nothing is selected, or (b) only the attached node is selected,
otherwise the base node will drag along as well and they will not become detached.

% To detach an attachable node from a base node (method 2)

An alternate way to detach a node from its base node is to right click on the attached node and
select Detach from the context menu. For bidirectional attachments (such as a lifeline attached to
an object), this is the only way to detach them.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 14 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 General Techniques 15

Comments

You can add comments anywhere within a type of UML diagram using a Note symbol. There’s
nothing to prevent you from simply adding text annotations throughout your diagram for a similar
purpose, but UML prefers that comments be contained within Note symbols. An added advantage
is that you can use an Anchor path to attach a Note to a particular diagram element or direct it to
any area of a diagram.
Comments can be simple descriptions that annotate your diagrams, or they can be actual
components of diagrams such as constraints on classes or associations, or extension points in use
cases.
When you want to associate a comment with a region of a diagram rather than with a specific

diagram element, you can use an anchor that ends with an anchor point as shown. An anchor point
is a small circle that indicates that the comment is anchored to a position in a diagram rather than
to a node or path. By default, an anchor point appear on an anchor when the terminating end it
unattached to a path or node. You can create such an anchor by entering connect mode, choosing
the anchor style, clicking on the note, then double clicking where the anchor should end.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 15 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
16 General Techniques

Containers
UML makes liberal use of containers, large outline shapes that group sections of diagrams.

NOTE: The term container comes from Pacestar UML Diagrammer rather than from UML and
represents a clear and prevalent concept throughout UML and other modeling diagram
notations.

Some containers such as frames and subject boundaries are specifically tasked for sectioning and
organizing a diagram. Others serve as enlarged representations of nodes such as states, classes,
components, and packages that are shown expanded in order to express internal implementation
or structural details. A superstate is typical of many UML containers.

An obvious way to create a container based on an existing node is to simply create a node
symbol, enlarge it, and add the internal diagram elements within. However, using the specialized
container style for the corresponding node offers many advantages:
z Containers are transparent so that you need not be concerned whether the internals are in front
of or behind the container.
z Containers are selectable only by clicking on their outline (or text area if one is present). This
makes it much easier to work on the encompassed diagram details without being concerned
about the container. For example you won’t accidentally select the container when you click
on what seems to be open space at the more detailed level.
z Most containers are configured to be created by a lasso method. Instead of stamping out a
fixed size initial shape, you click where you want the upper left hand corner of the container
and drag the mouse to establish its bounds, ideal for sectioning off a portion of a diagram.
z Containers have other minor properties that are designed for the convenience of their intended
purpose such as how they resize, snap to the grid, and support attachments and path
connections.
Containers are available from the Containers drop menu which is located within the Nodes drop
menu on the style bar on the left hand side of the screen.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 16 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 General Techniques 17

Frames

Frames are a new type of container introduced in UML 2.0 that are used in UML diagrams to
show groupings and constructs that apply to portions of a diagram. You can draw a frame around
a portion of a diagram that includes any number of objects or even other frames.

Frames have specific uses within diagrams such as combined fragments in sequence diagrams but
they are available in all UML diagrams. One common use is to bound any diagram or diagram
fragment to identify the type of diagram, such as class, component, communication, sequence and
so on.

% Creating Frames
Create a frame by choosing its icon from the Containers submenu located within the Node Styles
drop-down menu in the style bar. Position it by clicking where you want the upper right corner
and dragging to define the size of the frame. Then switch to text mode and edit the description
text.

% Selecting Frames
To select a frame, click on the outline or in the description area. Drag the frame from the same
areas. Like other containers, the frame is transparent so that you can see diagram element it
surrounds even if the frame is on top of them. Therefore, you cannot select it or drag it by clicking
on its interior as you would with ordinary nodes.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 17 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
18 General Techniques

% Resizing Frames
When you add text to a frame, the description area will expand as needed to contain the text. You
can also reshape this area by selecting the frame and dragging the yellow reshape handle.

Path Labels
Path labels occur frequently in UML diagrams. They are used to add guard conditions, name
associations, identify constraints, express relationships, and so on. Pacestar UML Diagrammer
supports several different path label types which remain properly positioned when the path or
attached nodes move around.

In-line vs. Lateral Path Labels

The most common types of path labels are in-line and lateral path labels. Both serve the same
purpose of labeling the path itself and both can be placed anywhere along the length of a path.
The only difference is in appearance. An in-line label is drawn directly within the path
interrupting the path, whereas a lateral label is drawn alongside the path.

UML allows for either style when a path requires a label. You are free to choose the method that
is most clear and fits the constraints of your diagram.
When in text mode the presence of a dotted rectangle indicates what type of label will be created
when you click at the current cursor position. If no rectangle appears, the label will not be
attached to a path or other object. If the rectangle appears directly over the path, the label will be
an in-line path label. And if it appears alongside the path, it will be a lateral label.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 18 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 General Techniques 19

Flow Labels
Flow labels provide a way to label the intersection of a path with a node. Unlike in-line and lateral
path labels, flow labels are concerned with both the path and the attached node and are
automatically positioned in the “corner” where the two meet regardless of angle or direction.

Common uses for flow labels include multiplicities and role names on associations. Note that a
single path can have up to four flow labels, one on each side of the connection of each end of the
path.
When in text mode the presence of a dotted rectangle indicates what type of label will be created
when you click at the current cursor position. If the rectangle appears in the corner of the
intersection of a path and a node, the label will be a flow label. You can tell it will be a flow label
rather than a lateral label both because it is fixed at the very end of the path and by the presence of
a small dotted line that appears between the rectangle and the point where the path meets the
node.

Off Center Path Connections

The simplest diagrams are constructed from nodes connected by single paths from one node to
another. With small numbers of simple connections, paths that attach toward the center points of
the nodes are the most basic and easiest to manage. Some diagrams such as activity and state
diagrams lend themselves almost exclusively to this type of construction because each node has a
limited number of entering and exiting paths. Class diagrams, on the other hand, lend themselves
to similar construction only when they are very simple. For example, here are two very basic class
diagrams using centered and off center path connections.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 19 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
20 General Techniques

There are two ways to control which of these connection models are used to configure new path
connections. You will find this information under Attach Modes in the main user guide, but we’ll
summarize it here as well for convenience.
If most of your connections will be centered with an occasional off center connection, you can
switch back and forth simply by holding down the SHIFT key as you make the connection. For
example, if you are attaching a path between two nodes and wish it to originate from the point on
the first node where you click (as opposed to the center of the node), you simply hold the SHIFT
key as you click to position the start of the path on the first node. This method of connecting a
path is called Click Point. Note that while in connect mode, the status bar will remind you that
holding SHIFT will result in Click Point attachments. Also note that you can route the path from
the origin point on the first node to the terminating point on the second node by clicking at any
number of intermediate points on the diagram in between.
You can also switch attach modes so that holding down the SHIFT key is no longer necessary. If
your diagram contains many off center attachments, this is the more practical way to proceed. In
fact, switching the attach mode to Click Point is often the first step in beginning a non-trivial
class diagram. The attach mode can be changed by selecting one of the modes listed under Attach
which is available in the Paths menu. Once in Click Point attach mode, SHIFT switches back to
centers on a case-by-case basis.

Path Trees
Some UML diagrams require more than simple node-to-node paths. You can create any
conceivable circuit of paths by connecting portions of paths to one another. The most common
arrangement is a “tree”. A tree minimizes the number of paths that connect to a common node. A
good example is an inheritance relationship which consists of a large number of classes that
derive from a base class, each depicted by a generalization path to the base class node.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 20 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
General Techniques 21

Creating Path Trees

It’s easiest to create a tree when using the snap grid so that the paths line up with minimal effort.
If the snap grid is not enabled, or if the arrangement of the nodes is not just right beforehand, you
can always go back and adjust the paths for better alignment.

1. Choose your connector style and enter connect mode. In this example we use a generalization
path. Create the first leg of the tree by clicking at point A (anywhere on the lower class node)
and then at point B (anywhere on the upper class node).
2. Create the left branch of the tree by clicking at point C (anywhere on the left class node), then
at a point D of your choosing directly above the class, and terminate the branch by clicking on
point E directly to the right of D and on the path you created in the last step. You can tell that
the cursor is over the vertical path when the arrowhead of the cursor changes from white to
black.
3. Repeat the last step from points F and G, and back to E. If any of the connections did not turn
out straight, go back to select mode (ESC) and drag the junctions that are now located at points
D, E, and G until they appear correctly.
You can see that a simple tree like the one shown is just one example of a limitless arrangement of
paths that you can design.

Manipulating Path Trees

A tree or any other structure that involves a number of nodes and paths can be tricky to edit,
move, add to, or delete from. The main user guide describes many techniques that might be
required to accomplish a wide range of editing tasks. The most broadly useful technique is simply
selecting the correct subset of objects using lasso selection. Selecting the proper set of objects to
move, copy, or delete will help to avoid unnecessary clean-up steps.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 21 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
22 General Techniques

1. In this example, we enlarge the tree vertically by selecting the bottom portion and dragging it
downward in a single step.

2. In this example, we move the right branch to the right as if making room for a new class.

3. In this example, we move the connecting paths upward. When dragging paths like this, we
have to make sure to grab the selected path either by the horizontal portion of one of the
selected paths, or by the small red outline around the selected junctions - NOT by the red
handle on the junctions which will cause the movement only of the one junction.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 22 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 General Techniques 23

Node Symbols Containing Internal Icons

Several node symbols contain internal icons (usually in the upper right or lower right corners) to
add clarity or to further specify their role in the diagram. A number of these symbols are shown
here:

The icons are defined as part of the node styles. The line width and color they are drawn with
matches the line width of the border of the style and cannot be modified independently. The
spacing between the icon and the edges of the shape matches the text border and can be controlled
(currently only per-diagram) by adjusting the text margin spacing located in the Diagram
Properties dialog box.
Also note that in most cases the internal icon overlaps the text area of the symbol. This means that
your text may collide with the symbol. You can avoid this manually by adjusting the size of the
symbol, the text justification, or padding the text with extra spaces.

Extensions and Nonstandard Symbols

The tool does not force you to adhere to the strict definition of UML. Instead it allows you to
extend the language as you see fit for your own needs. We leave it up to you and your best
judgement to determine whether to use custom extensions and notation.
UML does not preclude adding symbols that are not a part of the UML specification to your UML
diagrams for the purpose of enhancing or adorning your diagrams. Each of our UML diagram
templates include a category named “Nonstandard” that contains a number of shapes that may be
useful in your diagrams but are not included in UML and have no defined usage. They are
included for you to use (or not use) at your own discretion.
The Nonstandard category of symbols are a subset of the larger collection of symbols included
with the product in the symbol galleries. The symbol galleries contain a variety of addition shapes
and symbols such as obsolete symbols from older UML standards, more nonstandard UML
symbols, flowcharting symbols, and other miscellaneous shapes. To add a symbol or icon from
the symbol galleries use the Insert Node/Icon feature from the Nodes menu (also available by
default as a toolbar button).
Similarly, you may on occasion want to add imported clipart or graphics to your diagrams. This is
fully supported by the tool and is commonly practiced with UML. In Use Case diagrams for
example, many designers prefer to substitute clipart or icons for actors to increase clarity or add

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 23 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
24 General Techniques

emphasis. To insert a graphic from a file, use the Insert Graphic feature from the Nodes menu
(also available by default as a toolbar button).

Miscellaneous Drawing
You may find the need to draw diagram elements that are not included in the UML diagram
templates. There are several reasons why this could happen:
z You require a symbol or construct that is not supported by the tool.
z You need to add a simple dividing line, arrow, box, etc. that is not a standard part of UML
such as to annotate or adorn your diagram.
z You decide to implement your own extensions.

We have included some generic shapes and line just for this purpose.
Do not use diagram elements for general drawing purposes
We recommend that you resist the temptation to use a UML diagram element for a purpose for
which it is not intended just because it is the proper shape or line type. If you need a generic box
for example to add the author’s name at the bottom of a page, do NOT use a Class/Object symbol.
If you need a simple line to divide your diagram into two parts, do NOT use an Association path,
and so on. The UML diagram elements have defined purposes and it’s best to use them only for
these purposes. Adding extraneous diagram elements could hinder consistency checking, confuse
third party tools that attempt to interpret your files, and interfere with conversions when
upgrading to new releases of the tool or new versions of the UML specification.
Instead, we have included some basic generic symbols and lines especially for this reason. In the
previous section we discussed how to add nonstandard shapes both from the Nonstandard Node
Style drop menu, and from the Symbol Galleries via the Insert Node/Icon feature. We have also
included a category of path styles for generic drawing. You will find these in the Nonstandard
drop menu category by clicking on the Paths button in the style bar at the left of the screen. If you
need to add a line to your diagram, pick one of these styles and draw with it. (Remember that a
line does not need to start or end on a node, you can start it anywhere and end anywhere else by
double-clicking.) Once you’ve created the line, you can select it and use the toolbar controls or
the Path Properties to change its color, thickness, line pattern and so on however you like.
For creating compound symbols or for adding new lines to existing symbols, you can try using
the Group feature. A group has limited functionality but it can be a helpful way to create a quasi-
custom symbol that you will subsequently copy within your diagram. Note that the group feature
is not intended for this purpose so the grouped elements will not behave like a new shape or style
- for example you cannot use a group to define the shape of a node style and you cannot add a
group to a style bar button for easy replication. One example where grouping could be useful is in
creating an unusual-shaped partition buy grouping lines, bookstand even labels into a single
structure that can then be copied, moved, duplicated, and resized as a whole.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 24 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
2 ACTIVITY DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Activity Diagrams.

Sample Activity Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 25 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
26 Activity Diagrams

Activity Diagram Style Usage Tables

ACTIVITY DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Flow flow Flows (equivalently called edges) connect actions to show

edge control flow.

Object Flow object flow An input or output to or from an Object Flow State symbol.
Flow or Object Flow path styles are used for this purpose
seemingly interchangeably. Object Flow paths provide addi-
tional clarity.

ACTIVITY DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Action action (activity) An action state consisting of a single activity that runs to
completion.

Subactivity subactivity An action that is actually an entire subactivity that is or can

be expanded and decomposed.

Object Flow object flow state A key data object state shown being affected by actions.
State

Send Event send event A signal that is sent asynchronously to a target. Use this
symbol when the target is not shown, or when a compli-
mentary receive event is shown to the right.

Receive Event receive event A signal that is received from a target. Use this symbol
when the source is not shown, or when a complimentary
send event is shown to the right.

Send Event2 send event A signal that is sent asynchronously to a target. Use this
(opposite direc- symbol when a complimentary receive event is shown to
tion) the left.

Receive Event2 receive event A signal that is received from a target. Use this symbol
(opposite direc- when a complimentary send event is shown to the left.
tion)

Fork/Join Horz fork, A horizontal fork or join symbol (used interchangeably). A

join fork has one input on one side and multiple outputs on the
other side. A join always accompanies a fork and has mul-
tiple inputs on one side and a single output on the other.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 26 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Activity Diagrams 27

ACTIVITY DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Fork/Join Vert fork, A vertical fork or join symbol (used interchangeably). A

join fork has one input on one side and multiple outputs on the
other side. A join always accompanies a fork and has mul-
tiple inputs on one side and a single output on the other.

Branch/Merge branch, Decision branch point where the guard conditions ade-
merge quately describe flow alternatives. Use the same symbol
as a merge point.

Decision decision This larger decision symbol is like a branch symbol but it
can also contain text to help describe the branch condi-
tions, reducing the complexity of the guard conditions on
the outgoing flows.

Initial State initial state Initial state. This is the starting point of the activity.

Final State final state Final state. This is the ending point of the activity.

Expansion expansion region A container for defining an action state expansion region.
Region The interior contains the expansion of the action state.

Pin pin Pins represent input or output data parameters for an

activity/action. A pin can be attached to an action or
expansion region.

Listbox Pin Vert listbox pin Also known as an expansion node, a listbox pin repre-
sents a list of input or output data parameters for an activ-
ity/action. A listbox pin can be attached to an action or
expansion region.

Listbox Pin Horz listbox pin A horizontal version of the listbox pin for attaching to the
top and bottom sides of actions or expansion regions.

Exception exception param- This small symbol is placed near an output pin to repre-
Parameter eter marker sent an exception parameter output that flows to the next
action immediately

Time Signal time signal A time signal (or accept) triggered by passage of time (or
triggered to wait.)

Flow Final flow final An end to a flow that does not terminate the activity.

Connector connector A shorthand symbol for continuing a path at another loca-

tion. Connectors are always present in identically labeled
pairs.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 27 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
28 Activity Diagrams

Forks and Joins

Forks and joins split and re-join flow control for synchronizing multiple concurrent flow paths.
The same symbol is used for both forks and flows, and a vertical and horizontal version is
available.

Forks and joins always occur in pairs. A fork usually has a single flow entering one side and
multiple flows exiting the other. The complimentary join usually has the same number of flows
entering one side and a single flow exiting the other side.
To create a fork or join, add a “Fork/Join Vert” or “Fork/Join Horz” node to your diagram. Then
connect flows to and from the incoming and outgoing actions. Once a connector is attached, you
can select the connector and slide its connecting end to any point on the fork/join to achieve the
spacing you prefer. You can also resize the fork/join by selecting it and dragging the bright red
handles on either edge. Note that you can only lengthen or shorten the fork/join in this manner. If
you need a thinner or thicker fork/join, select the symbol, right click and choose properties, the
edit the height and/or width to change the size.

Object Flows and Object Flow States

Object flow states in activity diagrams are simply objects like those that appear in object
diagrams (instances of classes). They represent data that is read and modified by actions.

UML references vary as to whether normal flows should be used to connect actions and object
flow states, or whether a special dashed Object Flow path should be used. We included the Object
Flow path style for this use, but you are free to use ordinary flows.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 28 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Activity Diagrams 29

Conditional Branches and Decisions

Conditions in activity diagrams can be represented in several ways.

Branch node
Use a simple branch node to show that the flow branches in different paths based on conditions
that can be easily described in the guard conditions of the outgoing flows. For example, a branch
could have two outgoing flows with the guard conditions “[daytime]” and “[nighttime]”.
Decision node
When the guard conditions alone are insufficient for describing the condition, use the large
Decision symbol so that you can use text to describe the condition more clearly. A decision node
is analogous to those found in flow charts. For example, rather than using a simple branch node
with elaborate guard conditions such as “[date of last internal audit more than a year ago]” and
“[date of last internal audit less than a year ago]”, you could instead use a Decision with the text
“date of last internal audit” and guard conditions “[more than a year]” and “[less than a year]”.
Action node
An action node can be used in place of a decision node when the branch of flows is the result of
some action described in the action node. For example, an action node might contain the text
“Calculate the rate of return.” and the guard conditions could be “[ROI > 5%]” and “[ROI <=
5%]”.

Pins
Pins represent data flows such as input and output parameters. Pins typically attach to the sides of
actions to show data inputs and outputs, but the can also occur outside actions as if they were
miniature versions of object flow nodes. They can also appear on the edges of expansion regions

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 29 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
30 Activity Diagrams

and decomposed actions to show how the action or expansion region interacts with data. In the

latter case, the pins appear directly overlapping the outline of the node.
Pins are attachable nodes that attach to the edges of actions, or overlapping the edges of
expansion regions. Like all attachable nodes, you create the base node first (in this case the action
or expansion region), then create the attachable node (in this case the pin) and snap it into place
on the base node where it attaches for the purposes or moving, copying, and so on (see
Attachable Nodes on page 12).

% To create a pin
1. Select the Pin node style. The cursor will take the shape of the pin.
2. If creating a free pin (not attached to a node), simply click on the diagram. If creating a pin on
an action, move the cursor until it is aside the outer edge of the action and it snaps into place. If
creating a pin on an expansion region or other action container, move the cursor until it is
directly over one of the edges and it snaps into place.
3. Click to create the pin. The pin will automatically cling to the action or expansion region if you
clicked near a node of the proper type.

% Manipulating a node that has attached pins

Whenever you drag, copy, duplicate, or delete a node (usually an action) that has one or more
pins, the pins will move with the action, be copied with it, or be deleted with it.

% Labeling a pin
The best way to label a pin is to attach a node label to it. In the above example, you could right
click on the pin and choose Add Node Label, Above Left. Use text mode to change the text of the
label to the name you want. You may also want to reposition the label relative to the pin by
dragging it. The label will remain attached to the pin in the same way the pin is attached to the
action. You can un attach it later if you like by right clicking on the label and choosing Detach
Label.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 30 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Activity Diagrams 31

% Detaching a Pin
To detach a pin from an action, simply drag it away from the action node. Alternately, right click
on the pin and select the Detach command.

% Repositioning a pin
You can drag a pin to reposition it on the action or detach it. Dragging to reposition the pin works
similar to creating the pin. You can choose to move it free of the action, attach it to a side of an
action, or attach it to an expansion region.

Listbox Pins
Listbox pins are synonymously called expansion nodes. We’ve chosen the term listbox pin
because it is more easily remembered and associated with pins. A listbox pin is similar to a
standard pin except that it typically identifies a range of data for which the expansion region will
act upon. For example, a listbox pin could represent a customer database or list of names and the
expansion region could represent a complex activity that is applied to the data set (or manipulates
it). The key representation made by having a listbox pin as opposed to a standard pin is that the
action, subactivity, or expansion region applies iteratively.

Creating, manipulating, labeling, and repositioning listbox pins work similar to with standard pins
(see previous section). The only difference is that listbox pins can be vertical or horizontal and
therefore should be attached to the corresponding sides of node.
Listbox pins are attachable nodes that attach to the edges of actions, or overlapping the edges of
expansion regions. Like all attachable nodes, you create the base node first (in this case the action
or expansion region), then create the attachable node (in this case the listbox pin) and snap it into
place on the base node where it attaches for the purposes or moving, copying, and so on (see
Attachable Nodes on page 12).

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 31 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
32 Activity Diagrams

Connectors
Connectors are small circles used to replace sections of paths. Use connectors anytime they
increase the readability of a diagram or simplify the maintenance of a complex routing of paths.
Connectors are also important for continuing a path from one page to another or one diagram to
another.

UML allows the use of connectors to replace just about any path. Therefore you will find the
connector symbol available is all UML diagram templates located in the General symbols under
Nodes. Simply add connectors to your diagram, one where the path breaks, and the other where
the path continues. Label each connector with an identical letter or number. Then connect up the
paths.

Exception Parameters
Exception parameters are a special case of an output pin which is marked by a small triangle. The

distinguishing characteristic of an exception parameter is that control flow occurs immediately

when the exception occurs whereas flow from other output pins always occurs at the same time
(when the action is entirely complete and all output is prepared).
The tool includes a symbol for an exception parameter that you can use to annotate an output pin.
The symbol does not automatically attach to the pin so you will need to select it along with the
action before dragging to keep them together.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 32 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Activity Diagrams 33

Interruptible Activity Regions

An interruptible activity region is marked by a dashed rounded rectangle (a container). Use the
node style “Interruptible Region” to enclose the interruptible portion of the activity diagram. You
can use a jagged flow path to indicate the flow that is followed when the interrupt occurs within
the region.

The jagged flow is not a special flow style, but rather a flow that is created in three segments. For
example, select the Path path style, click at A, click at B, click at C, then double click at D to
create the above jagged flow symbol.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 33 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
34 Activity Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 34 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
3 CLASS/OBJECT DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Class/Object
Diagrams. Techniques described elsewhere for handling messages, ports, interfaces, packages, and
partitions are also useful when working with class diagrams.

Sample Class Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 35 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
36 Class/Object Diagrams

Class/Object Diagram Style Usage Tables

CLASS/OBJECT DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Association association A relationship between classes or objects. An association

without arrows shows either bidirectional navigability, or
simply does not show the navigability.

Directional directional A association between classes that is navigable in one

Association association direction.

Bidirectional bidirectional A association between classes that is navigable in both

Association association directions.

unnavigable Although no styles are pre-defined with the “X” terminator

association that denotes unnavigability, you can add it to the end of any
(one direction) association (right click on the end) to show explicitly that the
association if not navigable in one direction.

unnavigable Although no styles are pre-defined with the “X” terminator

association that denotes unnavigability, you can add it to the end of any
(both directions) association (right click on the end) to show explicitly that the
association if not navigable in one direction.

Generalization generalization Shows a generalization relationship between actors (inherit-

ance). The arrow points toward the less general type.

Aggregation aggregation Expresses an “is a” relationship between classes, the com-

bination of a number of other classes to form a new class.

directional Although no style is pre-defined to show navigability on an

aggregation aggregation path, the notation is sometimes useful. You can
add the navigability terminator to the end by right clicking on
the end of the path and selecting the new terminator.

Composition composition Expresses a “has a” relationship between classes. One

class is included as a component of another class.

directional Although no style is pre-defined to show navigability on an

composition composition path, the notation is sometimes useful. You can
add the navigability terminator to the end by right clicking on
the end of the path and selecting the new terminator.

Containment containment Shows a containment relationship, a shorthand for showing

that a classifier contains others without placing them all
within the container package. Like composition but repre-
sents physical containment.
Dependency dependency Shows a dependency between classes or packages. The
arrow points to the entity that is depended upon.

Realization realization Shows an implementation relationship.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 36 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Class/Object Diagrams 37

CLASS/OBJECT DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Class class A class or object (class instance).

object

Class R2 class A class with a second compartment. The top compart-

object ment is generally used for the class name, and the
lower compartment for attributes.

Class R3 class A class with three compartments.The compartments are

object generally used for class name (top), attributes (middle),
and operations (bottom).

Active Class active class A class or object that actively directs other objects. In
active object earlier versions of UML an active class was shown with
a thick border instead.

Active Class R2 active class A class or object that actively directs other objects, with
active object a second compartment. The top compartment is gener-
ally used for the class name, and the lower compart-
ment for attributes.

Active Class R3 active class A class or object that actively directs other objects, with
active object three compartments. The compartments are generally
used for class name (top), attributes (middle), and oper-
ations (bottom).

Class Actor actor An actor class, same as a class with the “<<actor>>”
stereotype.

Class Actor R2 actor An actor class with two compartments.

Class Actor R3 actor An actor class with three compartments.

Class Boundary actor A boundary class, same as a class with the “<<bound-
ary>>” stereotype.

Class Boundary R2 actor A boundary class with two compartments.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 37 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
38 Class/Object Diagrams

CLASS/OBJECT DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Class Boundary R3 actor A boundary class with three compartments.

Class Control actor A control class, same as a class with the “<<control>>”
stereotype.

Class Control R2 actor A control class with two compartments.

Class Control R3 actor A control class with three compartments.

Class Entity actor An entity class, same as a class with the “<<entity>>”
stereotype.

Class Entity R2 actor An entity class with two compartments.

Class Entity R3 actor An entity class with three compartments.

Actor actor Shorthand notation for an actor class.

Control control Shorthand notation for a control class.

Entity entity Shorthand notation for an entity class.

Boundary boundary Shorthand notation for a boundary class.

N’ary Associator n’ary associator Joins multiple associations when more than two classes
are involved.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 38 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Class/Object Diagrams 39

CLASS/OBJECT DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Parameter List template A template is a parameterized model element con-

structed by affixing a parameter list node to class or col-
laboration.

Qualifier qualifier A qualifier. Qualifiers attach to classes. They contain

text to qualify an association (which connects to the
qualifier).

Collaboration collaboration General arrangement of objects that interact to imple-

ment behavior.

Messages, Shown elsewhere Also commonly used - described elsewhere

Packages,
Components,
Ports

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 39 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
40 Class/Object Diagrams

Associations
Associations are the most common type of paths found in Class and Object diagrams. They show
relationships between classes and objects such as generalization, aggregation, composition,
navigability, and so on.
At its simplest, an association is a solid line connecting two classes or objects. However the
association can be adorned by different types of terminators (arrowheads) and labels that
represent attributes of the association like its name, direction, roles, multiplicities, and
constraints.
We’ve provided styles that represent the most common types of associations “Association”,
“Directional Association”, “Bidirectional Association”, “Generalization”, “Aggregation”, and
“Composition”. However you can use any association style to define an association between
classes and alter it by changing terminators or adding labels to arrive at any possible association.

Association Navigability

Navigability of associations is represented by simple stick arrows. Navigability is not always

shown in a diagram. I this case no arrows are included. Although in diagrams where navigability
is shown, an association with out arrows is synonymous with bidirectional navigability.
Optionally a small X can be included as an association terminator to explicitly show that the
association is unidirectional.

In addition to the normal “Association” path style, we include the two common variations
“Directional Association” and “Bidirectional Association”. We don’t provide styles for all the
possible permutations such as those that involve the less frequently used unidirectional terminator
shown. To get a unidirectional indicator, create a directional indicator, then right click on the
endpoint without a terminator and chose the terminator for the “X”.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 40 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
Class/Object Diagrams 41

Association End Labels

Associations make liberal use of end labels to add information to the meaning of the association.
Rolenames, constraints, ordering, visibility, interface specifiers, and changeability are all denoted

by end labels (as are multiplicities described in the following section.)

Pacestar UML Diagrammer supports end labels with a feature called Flow Labels in which a label
can be created that attaches to both the node and one end of a path. In the case of the association
shown here, the rolename “role” is attached as a flow label at the point where the association
meets a class node. As a flow label, whenever the nodes and path move relative to one another,
the label will move to a suitable location as close as possible to the endpoint of the path but not
overlapping either the path or the node.
Creating a flow label is simple. Choose your label style. Then position the text caret cursor over
the position where a flow label would appear. This will be confirmed by a small dotted rectangle
with another dotted line leading to the path endpoint. If you click to create the label it will then be
a flow label. Each of the four locations where the path meets the nodes can have a flow label
(above and below on each end). See the main user guide or online help for more detailed
instructions on using flow labels.

Association Multiplicities
Multiplicities show how many of one class are associated with another, or how many are
permitted to be. Multiplicities can be added to associations just like end labels described in the
previous section.

However, because multiplicities are usually one of a very few varieties, we’ve added shortcut
styles with predefined text so that you can add a multiplicity without even having to do any
typing. The pre-defined common multiplicities are “*” (meaning any number, zero or more), “1”
(obviously meaning just one), “0..1” (meaning one if any), or “1..*” (meaning at least 1). To add
one of these multiplicities, simple click on the corresponding style bar button, and then click it
into place, being sure to let it snap into a flow label slot as described for end labels in the previous
section. You can of course, edit the text after creating the label if you require a different
multiplicity expression.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 41 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
42 Class/Object Diagrams

Association Names
Associations can also have an optional name. An association name is usually a verb or verb
phrase such as “hosts” or “is hosted by”. The association name can be implemented as an inline

path label in which the association breaks around the label, or more commonly as a lateral path
label as shown in which the name appears alongside the association (above or below).
Lateral path labels are discussed elsewhere and are described in detail in the main product user
guide. In short, you can name an association by selecting a label style. Then move the text caret
cursor over the association path until you see a dotted rectangle appear and click. The name label
will appear with the default text “name” which you will then edit with the proper name. Once
created in this way, the name will remain with the association if the nodes or the path moves.

Association Direction Indicators

Named associations may also include an optional association direction indicator. This appears as

a small solid arrowhead beside the association name. Often a name is implied to mean left to right
or top to bottom and a direction indicator is only used when this convention is contradicted.
Others prefer to always include a direction indicator.
To add a direction indicator to an association name, right click on the name and choose “Direction
Indicator” from the right click menu, then select “None”, “Forward”, or “Backward”. Notice that
directions such as right and left are not used because the association can be oriented in any
direction - though typically the name direction indicator is clearest if the association is vertical or
horizontal. Forward means from the point where the path was started to where it was terminated
when it was first created, so a forward direction could easily be right to left or left to right.
Note that Pacestar UML Diagrammer places direction indicators on vertical associations
following the name text. This is different than in most UML references where the arrow appears
before the text. The meaning however remains clear and the name remains predictably near the
path.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 42 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
Class/Object Diagrams 43

Association Constraints
Association constraints are simply documented relationships between two associations, usually

{xor} meaning exclusively one association or the other exists at a time.

To create an association constraint, select the Anchor path style and click on one association, then
the other. The anchor path will connect to both associations. Then create a lateral path label to
include the constraint text. The process for creating the lateral label is identical to that used to
create an association name, but do not as a direction indicator in this case.

Association Classes
An association class is another common construct in class diagrams in which a class is used to

further define an association. It’s easy to create by creating the association class (no special style
is required class/object will do). Then connect the association class to the association using an
Anchor path.

N’ary Associations
Most associations tend to be binary (between two classes or objects). UML also supports

associations between more than two and calls it an “n’ary association”. To represent an n’ary
association, create an N’ary Associator node near the center of the nodes, then connect each node

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 43 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
44 Class/Object Diagrams

to the N’ary Associator and add the appropriate end labels and multiplicities as needed. You can
name any or all of the association links (and add direction indicators if you like), or you can add a
node label to the n’ary associator instead to add a single name to the association if you prefer.

Qualifiers
Qualifiers attach to class symbols to add information to association paths.

A qualifier can attach to any side of a class as show. And a single class can have any number of
qualifiers.

An association that is attached to a qualifier can be annotated in all the usual ways. You can add
an association name, direction indicator, multiplicities, rolenames, constraints and so on. Labels
that generally attach to the ends of the association as flow symbols where the association meets
the class, should instead can be placed where the association meets the qualifier.

Qualifiers are attachable nodes that attach to the edges of classes. Like all attachable nodes, you
create the base node first (in this case the class), then create the attachable node (in this case the
qualifier) and snap it into place on the base node where it attaches for the purposes or moving,
copying, and so on (see Attachable Nodes on page 12).

Adding Qualifiers to Classes

Qualifiers automatically cling to the edge of any class/object symbol.

% To add a qualifier to a class

1. Select the qualifier symbol from the style bar.
2. Move the cursor near any edge of a class symbol until it snaps into place. It you prefer the
qualifier not to snap into place, hold the CTRL key.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 44 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Class/Object Diagrams 45

3. Click to create the qualifier attached in place to the class.

4. Complete the qualifier by using text mode to add and edit its text. Note that as the text requires
more space, the qualifier will expand intelligently based on its position relative to the class
symbol.
Once a qualifier is attached to the edge of a class, it will remain with the class if you move the
class, copy it, duplicate it, and so on. If you delete the class, all attached qualifiers will be deleted
automatically as well. If you prefer to delete the class but not the qualifiers, simply detach them
first.

Detaching Qualifiers from Classes

Drag the qualifier itself without selecting the class. If you move the qualifier away from the class,
it will become detached. You can also move it to a new position relative to the class by dragging
to a new location along the edge where it can snap into place. Alternately, right click on the
qualifier and select Detach.

Templates
Templates are constructed by attaching a parameter list box to the upper right hand corner of a
class node or collaboration node as shown.

Parameter lists are attachable nodes that attach to the upper right hand corners of classes and
collaborations. Like all attachable nodes, you create the base node first (in this case the class or
collaboration), then create the attachable node (in this case the parameter list) and snap it into
place on the base node where it attaches for the purposes or moving, copying, and so on (see
Attachable Nodes on page 12).

% To create a template by adding parameter list to a class or collaboration

1. Create a class or collaboration node.
2. Select the Parameter List node style.
3. Move the cursor over the upper right hand corner of a class or collaboration symbol until it
snaps into place. If you prefer it not to snap into place, hold the CTRL key.
4. Click to create the template, the parameter list node attached in place to the collaboration
symbol.
5. Use the text tool to add the actual parameter list text to the parameter list node.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 45 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
46 Class/Object Diagrams

Once a parameter list is attached to a class or collaboration, it will remain with it if you move the
class or collaboration, copy it, duplicate it, and so on. If you delete the class or collaboration, the
parameter list will be deleted automatically as well. If you prefer to delete the class or
collaboration but not the parameter list, simply detach it first.
To detach a parameter list from a class or collaboration, right click on the parameter list and select
Detach from the right click menu.
A parameter list cannot be repositioned while attached to a class or collaboration. To reposition it,
detach it and re-attach it at a new location.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 46 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
4 COMMUNICATION DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Communication
Diagrams. Communications Diagrams use similar notation to Class/Object Diagrams and require few
new techniques.

Sample Communication Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 47 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
48 Communication Diagrams

Message Style Usage Tables

MESSAGE PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Message SYNC synchronous A synchronous message / procedure call (originator

message waits for response for continuing).

Message ASYNC asynchronous An asynchronous message (originator continues rather

message than waiting for response).

Message REPLY reply A reply/return from a synchronous message. Reply

messages in UML 2 are the same as return messages
from UML 1. Some sources, notably [BIB] define the
reply as having a solid arrow. Some earlier revisions of
[SPEC] supported this interpretation (though ambigu-
ously), but [UD3], [SPEC], and [REF] all seem to sup-
port this as the proper symbol for reply.

Message LOST lost message A message sent to a receiver unknown or outside of

scope.

Message FOUND found message A message received from a sender unknown or outside
of scope.

Message DATA data flow Although not exactly messages, data flows are concep-
tually similar to messages and their behavior and
implementation are identical to other message types.

The above message path styles are point-to-point which means they are defined so that you create
a message by clicking on the source, then clicking on the destination. All of the above also have
stamp styles which you simply choose and then stamp onto your diagram where you want them.
The stamp styles are defined with the same style names as above plus a prefix indicating the
orientation of the message. For example, the stamp styles for the synchronous message are:

Message SYNC Left Points left with label on top

Message SYNC Right Points right with label on top
Message SYNC UpR Points up with label on right
Message SYNC UpL Points up with label on left
Message SYNC DownR Points down with label on right
Message SYNC DownL Points down with label on left

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 48 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Communication Diagrams 49

Messages
Messages are common in several types of UML diagrams, most prominently in Communication
Diagrams and Sequence Diagrams. Messages are represented as directed arrows (paths) often
with a text label that describes the message.

Messages can be attached at both ends to a source and destination as with most messages in
Sequence Diagrams. They can also be free floating as with most messages in Communications
diagrams. For convenience, Pacestar UML Diagrammer offers two ways to add messages to
diagrams corresponding to these two common usages. Messages can be created point-to-point or
they can be stamped onto a diagram.
Each message type has two pre-defined path styles. One path style is defined for creating point-
to-point messages and is available in the style bar. Stamp messages are available only in
Messages drop menu (in the main Paths drop menu on the style bar). Stamp message styles are
defined in a variety or orientations for quick access. You can use point-to-point and stamp
messages interchangeably discriminating only based on how you prefer to add them to your
diagram. Both methods result in identical messages that can late be edited in the same ways.

Creating Messages (Point-to-Point)

% To create a message (point-to-point method)

1. Click on the style bar button corresponding to a point-to-point message style. The cursor will
reflect the usual connect cursor, a stick pointer with a small arrow head.
2. Click on the source node.
3. Click on the destination node to complete the message arrow.
4. Use text mode to modify the default text and finalize the message.
5. If required, you can slide the label to adjust its position on the message path (left, right, above,
below and so on.)

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 49 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
50 Communication Diagrams

Creating Messages (Stamping)

% To create a message (stamp method)

1. Click on the Paths button in the style bar to drop down the Paths menu. Then select the
Messages submenu. This will display the complete message drop menu as shown. The first
column in the Messages drop menu has an icon for the point-to-point path styles (those that are
duplicated as style bar buttons). The remaining columns are stamp styles for each message in a
variety of orientations. Select one of the stamp message icons. The cursor will become a
complete representation of the stamp style including a default text label.
2. Click on your diagram to add a free floating message and repeat as desired.
3. Use text mode to modify the default text and finalize the message.
4. If required, you can slide the label to adjust its position on the message path (left, right, above,
below and so on.) You can also select the message and drag either end to change the length,
orientation, and connections as well.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 50 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Communication Diagrams 51

Labeling Messages
Message labels are simply lateral path labels. You can manipulate them like any other lateral path
labels, adding, deleting, sliding, and editing in all the usual ways. As with all lateral path labels,
they self-adjust to remain near the path as the path changes. While UML seems to prefer you to
use lateral path labels to label messages, you can just as properly use inline path labels instead.
The pre-defined path styles (point-to-point and stamp) automatically add lateral path labels to
message paths as you create them, but you may also substitute inline path labels on a case-by-
case basis. After you create a message with a label, you can simply slide the label over the path to
make it inline.

Looping Messages
Communication diagrams often require looping messages as shown.

% To add a looping message to an object

1. Select the path style from the style bar (point-to-point message style).
2. Choose a curvature style from the toolbar (straight, rounded, curved, or smoothed)
3. While holding down the SHIFT key click at (A), a any point near the edge of the object where
the message will begin.
4. Click at intermediate points (B) and (C) to continue the message loop.
5. While still holding SHIFT, click at a point (D) back on the object to complete the message.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 51 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
52 Communication Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 52 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
5 COMPONENT/DEPLOYMENT DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Component/
Deployment Diagrams.

Sample Component Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 53 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
54 Component/Deployment Diagrams

Sample Deployment Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 54 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Component/Deployment Diagrams 55

Component/Deployment Diagram Style Usage Tables

COMPONENT/DEPLOYMENT DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

To Interface None A path that connects from a component (or other classifier)
to an interface symbol (a Provided Interface or a Required
Interface node).

Provided provided An assembly connector for extending a provided interface

Interface interface (lollipop).
Assembly assembly

Required required An assembly connector for extending a required interface

Interface interface (socket).
Assembly assembly

COMPONENT/DEPLOYMENT DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Component component A modular implementation unit with well-defined interfaces.

Provided provided A named set of operations that characterize the behavior

Interface interface provided by an element.

Required required A named set of operations that characterize the behavior

Interface interface required by an element.

Port port A connection point for grouping logically related provided

and required interface.

Complex Port complex port A complex port (port with multiple interfaces) that is oriented
Vert vertically and attached to the top or bottom of a component
or other classifier.

Complex Port complex port A complex port (port with multiple interfaces) that is oriented
Horz horizontally and attached to the left or right of a component
or other classifier.

Node node A physical computation resource (computer, memory, server

etc.)

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 55 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
56 Component/Deployment Diagrams

COMPONENT/DEPLOYMENT DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Artifact artifact A tangible piece of information usually produced by a soft-

ware development process.

Interfaces
Interfaces depict defined points of interaction to classifiers such as classes, packages, and
components. UML 2.0 defines two type of interfaces, required interfaces and provided interfaces.
Each is diagrammed identically with different interface symbols. In the case of a required
interface, the interface symbol is a small circle which has been referred to as a lollipop for its
appearance. The required interface is represented by a half circle large enough to interlock with
the provided interface symbol. The combination of these two notations is also referred to as ball
and socket notation.

Creating Interfaces on Classifiers

Pacestar UML Diagrammer provides two different mechanisms for creating interfaces. The first
way that we cover here is convenient for attaching interfaces to classifiers. The second way will
be useful for connecting up interfaces and assembly connectors among multiple classifiers and
interfaces.

% To create an interface on a classifier

1. Select either the Required Interface or Provided Interface node style. In the example show we
selected the Provided Interface node style. The cursor will represent the shape of the interface
symbol. In the example, a small circle.
2. Move the cursor near a classifier that can have an interface (a class, package, or component).
3. Click to create the interface. A To Interface path will automatically connect the interface to the
classifier and the interface symbol will become attached to the classifier. If the edge of the

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 56 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
Component/Deployment Diagrams 57

classifier closest to where you place the interface has a Port, the interface will automatically
connect up to the closest port.

In either case the interface created in this manner is attached to the classifier and can be
conveniently manipulated along with the classifier. If you move the classifier, copy it, delete it,
and so on, all of the attached interfaces will act as if they were a part of the classifier.
For added convenience, when you add an interface to a classifier it receives a default text label.
You can manipulate the text label in the usual ways. However, if you type the text for the label
immediately, before moving the mouse from over the interface symbol and without leaving the
current mode, the text you type will replace the default label text for the interface you just created.
Hitting ESC to terminate text editing then returns you to create mode to create additional ports if
you choose.
Interfaces can be constructed in two different ways. The first way involves creating a “Provided

Interface” or a “Required Interface” node near an eligible symbol (class, package, or component).
This method is convenient for defining the interfaces for the class, package or component, and for
manipulating them later.
The second method is to use Provided Interface Assembly and Required Interface Assembly path
styles. Rather than using a node for the interface symbol, these path styles include terminators
that are indentical in appearance to the interface symbol nodes described above.
Both of the above configurations can be used interchangeably and interlock properly together. We
recommend that you use the node symbols for establishing the interfaces that are essentially part
of a class, package, or component. And us the path styles for connecting these interfaces to other
interfaces or remote classifiers.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 57 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
58 Component/Deployment Diagrams

% To connect a path to an interface belonging to a classifier

1. Select either the Required Interface Assembly path style or the Provided Interface Assembly
path styles.
2. Click on the originating symbol.
3. Click on intermediate points as necessary to establish the route of the path.
4. Click on the interface belonging to a classifier to terminate the path. Normally you will use a
Provided Interface Assembly path style to connect to a Required Interface node and vice versa.

Labeling Interfaces

Interfaces defined on classifiers often have a name. Add a name using the node label. Simply
right click on the interface symbol and select Add Node Symbol and choose a label position. You
can then use text mode to change the label text and use select mode to drag the position of the
label relative to the interface symbol.

Moving Interfaces
When you establish interfaces on a classifier using interface nodes (the recommended method
described in the previous section), you can adjust the interface symbol by simply dragging them.
The connection to the classifier will automatically adjust to the new position.

Assembly Connectors

Often you will see a complete interface assembly connector like the one show drawn on a
diagram but neither side (the ball or socket) is in the immediate vicinity of a classifier (class,
package, or component). In this case, both ‘sides’ of the assembly connector can be created by
Interface Assembly path styles that meet at a common point.

% To create an interface assembly (neither side part of a classifier)

1. Select either the Required Interface Assembly or the Provided Interface Assembly path style (as
shown in the example) by clicking on its button in the style bar.
2. Click on the source (usually a classifier), then click on any intermediate points needed to
define the route. In the example, click on (a) then (b).

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 58 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Component/Deployment Diagrams 59

3. Double click on the place in your diagram where the interface connector will appear. In the
example, double-click on (c).
4. Select the opposing Interface Assembly path style (in the example this is the Required Interface
Assembly which will form the complementary side of the assembly connector) by clicking on
its button in the style bar.
5. Route the second path to the same location created by the loose end above and click on the end
of the other path. In the example, click on (d), (e), then double-click back on (c).

Ports
Ports in UML2 appear as small squares that overlap the edges of component symbols (though
they can also be used on packages and classes). A port is a higher level concept than an interface
and represents a collection of related interfaces. Therefore, you can attach interfaces to ports
instead of attaching them directly to classifiers.

Ports are attachable nodes that attach to the edges of components (and classes and packages). Like
all attachable nodes, you create the base node first (in this case the component, class, or package),

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 59 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
60 Component/Deployment Diagrams

then create the attachable node (in this case the port) and snap it into place on the base node
where it attaches for the purposes or moving, copying, and so on (see Attachable Nodes on page
12).

Creating Ports
Ports automatically cling to the edge of any component symbol (or other classifier symbol such as
a class or package symbol). Most ports are public and externally visible, however a port can also
be attached just inside the classifier boundary to represent a private (hidden) port.

% To add a port to a component

1. Select the port symbol from the style bar.
2. Move the cursor over any edge of a component symbol until it snaps into place. It you prefer
the port not to snap into place, hold the CTRL key.
3. Click to create the port attached in place to the component.
Once a port is attached over the edge of a component, it will remain with the component if you
move the component, copy it, duplicate it, and so on. If you delete the component, all attached
ports will be deleted automatically as well. If you prefer to delete the component but not the ports,
simply detach them first.

% To detach a port from a component

Drag the port itself without selecting the component. If you move the port away from the
component, it will become detached. You can also move it to a new position relative to the
component by dragging to a new location along the edge where it can snap into place.

Labeling Ports
The best way to label a port is to attach a node label to it. In the above example, you could right
click on the port and choose Add Node Label, Above Left. Use text mode to change the text of
the label to the name you want. You may also want to reposition the label relative to the port by
dragging it. The label will remain attached to the port in the same way the port is attached to the
component. You can unattach it later if you like by right clicking on the label and choosing
Detach Label.

Attaching Interfaces to Ports

See previous section regarding creating interfaces.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 60 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
Component/Deployment Diagrams 61

Repositioning Ports
You can drag a port to reposition it on the component or detach it. Dragging to reposition the port
works similar to creating the port. You can choose to move it free of the component or attach to a
different component or a different position on the same component.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 61 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
62 Component/Deployment Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 62 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
6 COMPOSITE STRUCTURE DIAGRAMS
Composite Structure Diagrams require no unique techniques, although they make significant use of
containers and attachable nodes.

Sample Composite Structure Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 63 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
64 Composite Structure Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 64 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
7 INTERACTION OVERVIEW DIAGRAMS
Interaction Overview Diagrams require no unique techniques beyond those used for Activity
Diagrams and Sequence Diagrams.

Sample Interaction Overview Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 65 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
66 Interaction Overview Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 66 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
8 PACKAGE DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Package Diagrams.

Sample Package Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 67 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
68 Package Diagrams

Package Diagram Style Usage Tables

PACKAGE DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Dependency dependency Shows a dependency between packages. The arrow points

to the package that is depended upon.

Realization realization Shows an implementation relationship.

Generalization generalization Shows a generalization relationship between actors (inher-

itance). The arrow points toward the less general type.

Containment containment Shows a containment relationship, a shorthand for showing

that a package contains others without placing them all
within the container package. Use like generalization.

PACKAGE DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Package package A common use case symbol.

Package model A package model.

Model2

Package subsystem An external person, process, or thing that interacts with a

Subsystem2 system.

Package C2 package A package symbol with two columns.

Package R2 package A package symbol with two rows.

Package C2R2a package A package symbol with two columns, one having two
rows.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 68 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Package Diagrams 69

Containments
Containment is simply the collection of a number of packages within a broader package. UML
provides a special symbol for expressing containment that saves diagram space and is easier to
manage and extend. Use the containment path style as shown similar to how generalization is
expressed. See the Trees section for tips on working with tree path configurations.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 69 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
70 Package Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 70 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
9 SEQUENCE DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Sequence Diagrams.

Sample Sequence Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 71 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
72 Sequence Diagrams

Sequence Diagram Style Usage Tables

SEQUENCE DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Message SYNC synchronous A synchronous message / procedure call (originator waits

message for response for continuing).

Message asynchronous An asynchronous message (originator continues rather than

ASYNC message waiting for response).

Message reply A reply/return from a synchronous message. Reply mes-

REPLY sages in UML 2.0 are they same as return messages in ear-
lier versions of UML. Some sources, notably [BIB] define the
reply as having a solid arrow (the symbol shown here is
then used for object creation). Some earlier revisions of
[SPEC] supported this interpretation (though ambiguously),
but [UD3], [SPEC], and [REF] all seem to support this as the
proper symbol for reply.

Message LOST lost message A message sent to a receiver unknown or outside of scope.

Message found message A message received from a sender unknown or outside of

FOUND scope.

Measure none Designates a specific time interval.

SEQUENCE DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Lifeline lifeline An object lifeline. Note that the Lifeline style is a node style
rather than a path style despite that it appears as a vertical
dashed line.

Activation execution The rectangle is technically called an execution specifica-

specification, tion (also a focus of control [BIB]) with it’s start (top) defined
focus of control as it’s activation. It can be left white or filled gray. It repre-
sents a defined portion of an object’s lifetime, often repre-
senting the execution of a procedure from activation until
returning a message. An activation attaches to a lifeline. It
can also be used in place of a lifeline.

Object object An object with a lifeline and/or activations that communi-

role cates with other objects via messages.
part

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 72 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Sequence Diagrams 73

SEQUENCE DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Active Object active object An object that actively directs other objects. In earlier ver-
sions of UML an active class was shown with a thick border
instead.

State state States are sometimes attached to activations (or lifelines) to

show the state of the objects or roles.

Frame frame A container for identifying related diagram elements. Used

for expressing fragments (iterations and conditions) within
sequence diagrams.

Combined combined frag- A container that separates conditional fragments of a dia-

Fragment2,... ment gram. Available in multiple numbers of compartments.

Terminate destruction event Marks the destruction of a lifeline or activation. The termi-
nate node is used for different purposes with different
names in different diagram types. In a state diagram the
same symbol denotes a terminate node. For simplicity, we
use this term for both cases, noting that the terminate node
is used to represent a destruction event in sequence dia-
grams.

Actor actor Shorthand notation for an actor class.

Lifelines
Lifelines form the foundation of a Sequence Diagram. They appear as vertical dashed lines
attached to objects.

NOTE: There is some discrepancy in terminology among reputable sources. [SPEC] describes
the rectangle as the lifeline, whereas [REF2] describes the rectangle as a role and the
dashed line itself as the lifeline of the role. We have chosen to stick with the previously
popularized terminology which seems to remain clear and accepted, and also better
suits our implementation. With regard to roles vs. objects, we have also chosen to stick
with object for the time being, as it is clearly an object symbol representing the broader
concept of a role. For the purposes of our implementation, this convention also permits
you to use any variation of a class/object node as the object or role.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 73 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
74 Sequence Diagrams

Activations and messages attach to lifelines to show sequences of messages and system events.

Lifelines are attachable nodes that attach to the bottom of objects (roles). Unlike other attachable
nodes, you can create either the object or the lifeline first, then create the other and snap it into
place on the base node where it attaches for the purposes or moving, copying, and so on (see
Attachable Nodes on page 12).

% To create a lifeline attached to an object

1. First create the object symbol in any of the usual ways.
2. Select the Lifeline symbol from the style bar. Note that the Lifeline style is a node style rather
than a path style! This may seem counter-intuitive, but it may be helpful to remember the
lifelines and activations are both node styles and behave similarly in your diagram as message
sources and destinations.
3. Move the cursor directly beneath the object symbol until the lifeline snaps into place. If you
prefer the lifeline not to snap into place, hold the CTRL key.
4. Click to create the lifeline. You can extend the lifeline either by dragging it downward before
releasing the mouse, or by selecting it afterwards and resizing it then.
Once a lifeline is attached beneath an object, it will remain with the object if you move the object,
copy it, duplicate it, and so on. If you delete the object, the lifeline and all attached activations
and messages will be deleted automatically. If you prefer to delete the object but not the lifeline,
simply detach the lifeline first.

Activations
This tool uses the term activation to refer to what UML 2 officially terms an execution
specification (also a focus of control [BIB]), a vertical rectangle attached to a lifeline to identify
defined portions of the life cycle. The UML 2.0 specification actually defines the term activation
specifically as the start (top) of a execution specification, but this broader usage is common and
well understood.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 74 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Sequence Diagrams 75

Activations typically attach to a lifeline, but they can also attach directly to an object or to another
activation, overlapping to represent nesting and recursion.

Activations are attachable nodes that attach to the bottom of objects. Unlike other attachable
nodes, you can create either the object or the activation first, then create the other and snap it into
place on the base node where it attaches for the purposes or moving, copying, and so on (see
Attachable Nodes on page 12).

% To create an activation attached to a lifeline (or another activation)

1. Select the Activation symbol from the style bar.
2. Move the cursor directly over a lifeline symbol (or another activation symbol) until it snaps
into place. If you prefer the activation not to snap into place, hold the CTRL key.
3. Click to create the activation. You can extend the activation either by dragging it downward
before releasing the mouse, or by selecting it afterwards and resizing it by dragging either of
the two bright red resize handles.

% To change the color or line style of an activation

An activation by default is a simple “white” rectangle. However, UML 2.0 states that they can
also be gray. You can make an activation gray by selecting it, choosing a gray fill color using the
toolbar fill button, and changing the outline color (also on the toolbar) to gray. There are also
cases where an activation is drawn with a dotted line. You can use the line pattern toolbar button
to make this change.
Once an activation is attached to a lifeline (or another activation), it will remain with the lifeline if
you move the lifeline, copy it, duplicate it, and so on. If you delete the lifeline, the activation and
all attached activations and messages will be deleted automatically. If you prefer to delete the
lifeline but not the activation, simply detach the activation first.

% To detach an activation from a lifeline (or another activation)

Drag the activation itself without selecting the lifeline. If you move the activation away from the
lifeline, it will become detached. Alternately, you can right-click on the activation and select the
Detach command.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 75 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
76 Sequence Diagrams

Terminate Nodes
A terminate node (sometimes called a stop or a destruction event) marks the end of an object’s
lifeline. It appears as a small ‘X’ located at the bottom edge of a lifeline or an activation.

Terminate nodes are attachable nodes that attach to the bottom edges of lifelines and activations.
Like all attachable nodes, you create the base node first (in this case the lifeline or activation),
then create the attachable node (in this case the terminate symbol) and snap it into place on the
base node where it attaches for the purposes or moving, copying, and so on (see Attachable
Nodes on page 12).

Messages
Messages in sequence diagrams flow from a source object’s lifeline (or activation), to a
destination object’s lifeline (or activation). The messages used in activations are rarely free
floating, and therefore you should generally create them using the point-to-point message path
styles. See the section on messages for further descriptions of point-to-point message path styles.

% To add messages to a sequence diagram

1. Select a point-to-point message path style.
The point-to-point message path styles are simply the message styles that appear in the Paths
section of the style bar on the left side of the screen, rather than the stamp versions of the
message path styles located in the Messages drop menu that are designed for stamping onto
diagrams rather than connecting from on place to another.
2. Click at the source of the message. This is most often a lifeline or an activation, although you
can start a message anywhere.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 76 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Sequence Diagrams 77

3. Click on the destination of the message to terminate it. The destination is most often a lifeline
or activation as well, though you can terminate a message anywhere (double click to terminate
when not over a node).
4. When you create a message, it will have a default text label attached as a lateral path label to
the center and above the message. Enter text mode to edit the label text appropriately.
5. If you prefer the message label located further left or right, or below the message simply drag it
where you like.

% To move messages on a lifeline or activation

1. Select the message revealing red handles on either end.

2. Select a handle and reposition it on a lifeline or activation.
3. Repeat for the opposite end if required.

% To slide a message up or down a lifeline or activation

1. Optionally select the message.
2. Grab the message by the line portion (not by the handles if it is selected).
3. Drag the message up or down.
A message dragged in this fashion can only be dragged to the top or bottom end of the lifeline or
activation it is attached to. Notice that you cannot drag the message by grabbing onto the label.

% To make room for additional messages on a lifeline or activation

The common procedure of making more space in a sequence diagram for the addition of new
messages is highly instructive. In the diagram below, we need to add some messages to the

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 77 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
78 Sequence Diagrams

lifelines beneath the location of the arrow. Before we do, we’ll need to shift a portion of the
diagram down to free up the needed space.

1. The first step is making room in this diagram is to extend the lifelines, lengthening them
downwards. Select each lifeline (individually) in turn, and drag it downward as shown. The
messages and activations attached to the lifeline will not be disturbed.

2. Next use lasso selection to select all of the messages below the arrow, and shift them
downward by grabbing and dragging any of the selected messages. Grab by the line, not by the
red select handles.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 78 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Sequence Diagrams 79

Recursion and Subactivations

Activations can be attached to other activations (overlapped) to show multiple simultaneous
instances of an object. This notation most often represents recursion, either direct in which an
object’s activation initiates additional subactivation(s) of itself, or indirect in which the object
passes control to another object that in turn creates the subactivation. Note that the term
subactivation is a fitting descriptive term but to our knowledge is not formally defined within the
UML specification.

Creating the overlapping subactivation is straight-forward. Simply snap a new activation symbol
into place over an existing activation.
In the case of direct recursion, you may need to create messages that connect from the activation
to the subactivation and back without interacting with any other objects. Obviously the usual
straight horizontal message path is inadequate for this purpose. Most depictions of such messages
show looped messages as in the example here.Creating messages of this sort requires a few extra
steps.

% To create a looped message between an activation and a subactivation

1. Select the message path style. Be sure to choose one of the point-to-point styles from the style
bar buttons rather than the stamp styles from the drop menu.
2. Select the curved path type from the curvature toolbar button. Shown below:.

3. Create the message by clicking at points A, B, C, and D as shown in the example above.
Create a looped message on a lifeline in the same way.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 79 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
80 Sequence Diagrams

Guard Conditions on Activations and Lifelines

Guard conditions can be added to activations and lifelines for use with combined fragments.

You can add a guard condition to an activation or a lifeline by selecting the guard condition label
style (or any other label style). Move it over the activation or lifeline until the outline box is
visible then click to create the label. The guard condition label style should have a white
background so it is visible atop the activation. If not, select the label and use the Fill toolbar
button to change the fill color to white.

States on Activations and Lifelines

State symbols can be attached to lifelines or activations as attachable nodes.

Like all attachable nodes, you create the base node first (in this case the lifeline or activation),
then create the attachable node (in this case the state) and snap it into place on the base node
where it attaches for the purposes or moving, copying, and so on (see Attachable Nodes on page
12).

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 80 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
10 STATE DIAGRAMS
This section describes diagramming techniques that are primarily applicable to State Diagrams.

Sample State Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 81 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
82 State Diagrams

State Diagram Style Usage Tables

STATE DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Transition transition A transition from one state to another or the same state. A
guard condition on the transition, usually represented by an
attached path label, indicates the event that caused the
transition.

STATE DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

State state A condition in which each event has a consistent

defined response.

State R2 state A state with a name compartment and a body compart-

ment.

State R2 state A state with added detail not shown.

HidDecomp

State R3 state A state with a name compartment and two body com-
partments.

Name Tab name tab A name tab is affixed to the left side of the top edge of a
state symbol as an alternate method of naming a state.

Junction State junction state Chains transition segments into single run-to-comple-
tion transition.

Sync State synchronization A synchronizing pseudo state connecting forks/joins on

state a concurrency boundary.

Branch/Merge branch, Point where transitions merge asynchronously or

merge branch based on guard conditions.

Decision decision This larger decision symbol is like a branch symbol but
it can also contain text to help describe the branch con-
ditions, reducing the complexity of the guard conditions
on the outgoing flows.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 82 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 State Diagrams 83

STATE DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Initial State initial state Initial state. This is the starting point of the state
machine.

Final State final state Final state. This is the ending point of the state
machine.

History State history state Restores the previous condition of the composite state.

Deep History State deep history state Restores any past condition of the composite state.

Entry Point entry point An externally visible entry point that identifies an inter-
nal state as a target.

Exit Point exit point An externally visible entry point that identifies an inter-
nal state as a source.

Entry/Exit Point alternate entry/ An alternate notation for entry or exit points. The node
exit point symbol contains the name of the entry or exit point, and a tran-
sition arrow connects to or from the state.

Fork/Join Horz fork, A horizontal fork or join symbol (used interchangeably).

join A fork has one input on one side and multiple outputs
on the other side. A join accompanies a fork and typi-
cally has multiple inputs on one side and a single out-
put on the other.

Fork/Join Vert fork, A vertical fork or join symbol (used interchangeably). A

join fork has one input on one side and multiple outputs on
the other side. A join accompanies a fork and typically
has multiple inputs on one side and a single output on
the other.

Terminate terminate node Marks a terminate psuedostate. The terminate node is

also used in sequence diagrams to represent a
destruction event.

Container State state A container that encompasses the internals of a state,

such as a superstate.

Container State state with concur- A container that encompasses the internals of a state,
CB2,.. rency bound- such as a superstate, containing two compartments
ary(s) separated by a concurrency boundary. Available in mul-
tiple numbers of compartments.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 83 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
84 State Diagrams

Superstates

A superstate is a state that has internal substates and contains a diagram of its inner workings. An
obvious way to create a superstate is to create a normal state, enlarge it, and add the substate
diagram within. However, we strongly recommend that you instead create the superstate using a
State Container. A State Container (like any container) is more that a large state. For one thing it
is transparent so that you need not be concerned whether the internals are in front of it or behind
it. It is transparent so that it is selectable only by its outline or its text area (if any), making it
easier to work on the internals without accidentally selecting the container. Containers also have a
number of other minor behavioral differences designed for exactly this purpose.
The substate diagram within a superstate is simply a normal state diagram and requires no special
techniques.

Concurrency Boundaries

A concurrency boundary is a divider drawn within a superstate that separates two or more
portions of the state that operate concurrently. It is represented as a simple dashed line dividing
the superstate symbol. Although the tool does not explicitly support this notation as part of the
state symbol, you can easily add your own concurrency boundary anywhere within a superstate
symbol. To create the concurrency boundary, use a path of style Dashed Divider (obtain this by
clicking on the Paths button in the style bar and locating it in the Nonstandard category drop
menu).

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 84 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 State Diagrams 85

Entry and Exit Points

Entry and exit points are shown as small circles attached to the borders of a state symbol. The exit
point symbol also includes an “X” through it. Entry and exit points show entry and exit to and
from different internal substates.

Entry and exit point nodes are attachable nodes that attach to the edges of states. Like all
attachable nodes, you create the base node first (in this case the state), then create the attachable
node (in this case the entry or exit point) and snap it into place on the base node where it attaches
for the purposes or moving, copying, and so on (see Attachable Nodes on page 12).

% To add an entry or exit point to a state

1. Select the Entry Point or Exit Point node style.
2. Move the cursor over any edge of a state symbol (including state container symbols) until it
snaps into place. It you prefer it not to snap into place, hold the CTRL key.
3. Click to create the entry or exit point attached in place to the state symbol.
Once an entry point or exit point is attached over the edge of a state, it will remain with the state if
you move the state, copy it, duplicate it, and so on. If you delete the state, all attached entry points
and exit points will be deleted automatically as well. If you prefer to delete the state but not the
entry and exit points, simply detach them first.
You can move the entry or exit point to a new position on the state by dragging it along the edge
where it can snap into place.

% To detach an entry point or exit point from a state

Drag the entry or exit point symbol itself without selecting the state. If you move the entry or exit
point away from the state, it will become detached. Alternately, right click on the entry point or
exit point and select Detach from the menu.

Labeling Entry Points and Exit Points

The best way to label an entry point or an exit point is to attach a node label to it. You can add a
node label by right clicking on the entry or exit point and choosing Add Node Label. Use text
mode to change the text of the label to the name you want. You may also want to reposition the

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 85 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
86 State Diagrams

label relative to the entry or exit point by dragging it. The label will remain attached to the entry
or exit point in the same manner the entry or exit point is attached to the state. You can unattach it
later if you like by right clicking on the label and choosing Detach Label.

Alternate Notation for Entry and Exit Points

In addition to the standard entry and exit point notation, UML allows for an alternate notation in
which an entry point node or exit point node containing the name of the entry or exit point
connects to or from the state symbol.

Implementing this notation is straight forward and requires no new techniques.

Self-transitions
Self-transitions are common in state diagrams in which an event results in a transition to the same
state.

% To add a self-transition to a state

1. Select the path style from the style bar - usually a Transition.
2. Right-click on the state and select the Add Self Transition command to create a self-transition
of the current path style in the area of your click. In the example, the designer right-clicked on
the upper right area of the state to create a self-transition on this corner. Alternately, clicking on
any of the other corners, or in the middle of any side will create a self-transition there.

% To add a self-transition to a state manually

1. Select the path style from the style bar - usually a Transition.
2. Choose a curvature style from the toolbar (straight, rounded, curved, or smoothed).
3. While holding down the SHIFT key click on any point near the edge of the state to start the
loop.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 86 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 State Diagrams 87

4. Click on intermediate points to route the path away from the state and back.
5. While still holding the SHIFT key, click on the endpoint of the state where the self-transition
will terminate.
Note that self-transitions created manually may need to be manually re-adjusted if the size of the
state changes. Self-transitions created using the Add Self-Transition command will self-adjust
when the state resizes. Either type will move automatically with the state.

State Name Tabs

A name can be affixed to a state symbol using a rectangle along the top edge called a name tab.

Name tabs are attachable nodes that attach to the top edges of states. Like all attachable nodes,
you create the base node first (in this case the state), then create the attachable node (in this case
the name tab) and snap it into place on the base node where it attaches for the purposes or
moving, copying, and so on (see Attachable Nodes on page 12).

% To add a name tab to a state

1. Select the Name Tab node style.
2. Move the cursor over the top edge of a state symbol (including state container symbols) until it
snaps into place. It you prefer it not to snap into place, hold the CTRL key.
3. Click to create the name tab attached in place to the state symbol.
4. Use the text tool to add the name of the state to the tab.
Once a name tab is attached to a state, it will remain with the state if you move the state, copy it,
duplicate it, and so on. If you delete the state, the name tab will be deleted automatically as well.
If you prefer to delete the state but not the name tab, simply detach it first.
To detach a name tab from a state, right click on the name tab and select Detach from the right
click menu.
A name tab cannot be repositioned while attached to a state. To reposition it, detach it and re-
attach it at a new location.

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 87 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
88 State Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 88 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
11 USE CASE DIAGRAMS
This section describes diagramming techniques that are primarily applicable to Use Case Diagrams.
For added convenience, when you add an actor symbol to a diagram (or any iconic stereotype
symbol), it receives a default text label. You can manipulate the text label in the usual ways. However,
if you type the text for the label immediately, before moving the mouse from over the actor symbol
and without leaving the current mode, the text you type will replace the default label text. Hitting ESC
to terminate text editing then returns you to create mode to create additional actors if you choose.

Sample Use Case Diagram

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 89 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
90 Use Case Diagrams

Use Case Diagram Style Usage Tables

USE CASE DIAGRAM PATH STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Association association An association between an actor and a use case.

Relationship relationship Connects one use case to another to show a relationship

(includes or extends).

Generalization generalization Shows a generalization relationship between actors (inherit-

ance). The arrow points toward the less general type.

USE CASE DIAGRAM NODE STYLE USAGE

Style UML
Appearance Description
Name(s) Construct(s)

Use Case use case A common use case symbol.

Use Case Ext extended use An extended use case.

case

Actor actor An external person, process, or thing that interacts with a

system.

Actor System system actor A more generalized representation of an actor. Used to dis-
tinguish non-user actors (such as systems) from the stan-
dard actor symbol. Graphics and other symbols are
acceptable as well.

Subject subject boundary A rectangular container that shows the boundaries of the
Boundary subject (system). Any container can be used for this pur-
pose.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 90 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Use Case Diagrams 91

Relationships
Relationships between use cases are expressed as dashed paths of the style Relationship plus an
inline path label with a keyword (<<includes>> or <<extends>>).

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 91 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
92 Use Case Diagrams

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 92 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
12 UNSUPPORTED
Some UML notation is not specifically supported by features of this version of the software. In most
of these cases, the notation is minor or infrequently used, or it can be easily simulated using generic
drawing capabilities. The most important omissions will obviously be candidates for addition in future
releases.

Timing Diagrams
Timing diagrams are included as part of the UML 2.0 specification but are not directly supported by
this release of the software. While there is no timing diagram template or symbols and features to
conveniently support constructing timing diagrams, you can use the standard templates and the
generic symbols (lines and basic shapes) to create simple timing diagrams without too much difficulty.

Parameter Sets
Parameter set notation is not supported directly. Parameter set notation consists of rectangles that
encompass a set of pins. As a work-around you can add a generic rectangle for this purpose. It will just
lack the automatic attachment capabilities and will require greater effort to maintain.

Explicit Directions on Input and Output Pins

Input and output pins can be distinguished explicitly by adding small arrow heads to the pin symbols.
The software does not support this notation. You can of course use the convention of left=input
right=output, or you can add direction indication using labels.

General Ordering Path

The general ordering path style is not supported. It appears as a dashed line with an arrow in the
center (rather than at the end). If necessary, you can create such a path on a case-by-case basis using
flow symbols feature (see the main user guide).

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 93 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
94 Unsupported

Use Case Stereotype Symbol

Rather than the conventional oval use case symbol, UML allows for a notation consisting of a
standard classifier rectangle with a use case stereotype icon instead. This notation is not currently
supported.

Information Class Stereotype Symbol

The information class stereotype symbol is not supported (small solid triangle).

Custom Stereotype Symbols

The UML specification implies that you can create your own stereotype symbols for custom
purposes. While the tool does not support this directly, you can import clipart or draw using the
generic drawing styles, then combine the results using grouping functions. However, these
method will be cumbersome to maintain and are discouraged unless absolutely necessary.

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 94 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
CHAPTER
13 REFERENCES
The following are reference publications used in compiling this document and the features of the
software. The bracketed code preceding each is used within the text to cross reference the
publications.

[SPEC] UML 2.0 Superstructure Final Adopted specification

(formal/05-07-04, August 1995)
Object Management Group (http://www.omg.org)

[UD3] UML Distilled Third Edition

Martin Fowler, Addison-Wesley, 2004

[BIB] UML Bible

Tom Pender, Wiley Publishing, Inc., 2003

[REF2] The Unified Modeling Language Reference Manual, Second Edition

James Rumbaugh, Ivar Jacobson, and Grady Booch, Addison-Wesley, 2005

UML Diagramming Guide

EDGE Diagrammer User’s Guide

Page 95 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
96 References

UML Diagramming Guide

EDGE Diagrammer User’s Guide - Beta Version

Page 96 of 96
Last Modified December 28, 2005 5:01 pm
Filename: S:\UFLOW\UMLDOC\UML Diagramming Guide.fm
 Index 97

INDEX binary association 43

boundary 38
branch 27, 29, 82
branch node 29
A
action 26 C
action node 29 changeability 41
action state 26 class diagram 35
activation 72, 74, 75 class/object diagram 35
active class 37 classifier 60
active object 37, 73 collaboration 39
activity 26 combined fragment 73
activity diagram 25 comments 15
actor 38 complex port 55
aggregation 36 component 55, 60
alternate entry point symbol 83 component diagram 53
anchor 15, 43 composite structure diagram 63
anchor point 15 composition 36
angle brackets 11 concurrency boundary 84
assembly connector 55, 58 condition 29
association 36, 40 conditional branch 29
association class 43 connector 32
association constraint 43 constraint 41, 43, 44
association direction indicator 42 container 16
association end label 41 containment 36, 48, 68
association multiplicities 41 control 38
association names 42, 44 conventions (used in this guide) 9
association path 44, 51 Creating 60
association terminator 40
asynchronous message 48, 72 D
attachable node 12, 13, 31, 44, 45, 59, 74, data flow 29, 48
75, 76, 80, 85, 87 decision 27, 29, 82
decision node 29
B deep history state 83
ball and socket notation 56 dependency 36, 68
base node 12, 13 deployment diagram 53
bidirectional association 36 destruction event 73, 76, 83
bidirectional navigability 40

UML Diagramming Guide

98 Index

detach 14 guard conditions (on activations and life-

direction indicator 42, 44 lines) 80
directional aggregation 36 guillemets 11
directional association 36, 40
directional composition 36 H
drawing 24 hidden port 60
history state 83
E
edge 26 I
end label 41 initial state 27, 83
entity 38 inline path label 18, 42, 51
entry point 83, 85 intefaces 60
exception parameter 27, 32 interaction overview diagram 65
execution specification 74 interface 35, 47, 56
exit point 83, 85 interface assembly connector 58
exit points 85 interface label 58
expansion node 31 interface specifier 41
expansion region 27, 29, 30 internal icon 23
extensions 23 interruptible activity region 33

F J
final state 27, 83 jagged flow 33
flow 26 join 26, 27, 28, 83
flow final 27 junction state 82
flow label 19, 41
flow symbol 44 K
focus of control 72, 74 keys 9
fork 26, 27, 28, 83 keywords 11
found message 48, 72
frames 17 L
free floating messages 49 labeling interfaces 58
free pin 30 labeling messages 51
labeling ports 60, 85
G lateral path label 18, 42, 43, 51
general ordering path 93 lifeline 72, 73, 75
general techniques 11 listbox pin 27, 31
generalization 36, 68 lollipop interface 56

UML Diagramming Guide

 Index 99

looped message 51, 79 partitions 35, 47

lost message 48, 72 path label 18
path trees 19, 20
M Paths 49
merge 27, 82 pin 27, 29
message label 51 point-to-point messages 49, 76
messages 49 port 35, 47, 55, 57, 59, 60
point-to-point 49 private port 60
stamped 49, 50 provided interface 55, 56
messages (in sequence diagram) 76 provided interface assembly 55, 57
model 68
moving interfaces 58 Q
multiplicities 41, 44 qualifier 39, 44, 45

N R
n’ary association 43 realization 36, 68
n’ary associator 38 receive event 26
n’ary associator node 43 recursion 79
name tab 82, 87 reply 48, 72
navigability 40 required interface 55, 56
node 55 required interface assembly 55, 57
nonstandard symbols 23 role 41, 72, 73
note 15 rolename 41, 44

O S
object 72 self-transition 86
object diagram 35 send event 26
object flow 26, 28 stamped messages 49, 50
object flow node 29 state 73, 82
object flow state 26, 28 state container 84
state diagram 81
P stop 76
package 60, 68 subactivation 79
package diagram 67 subactivity 26
parameter list 45 subsystem 68
parameter sets 93 superstate 84
part 72 symbols (used in this guide) 9

UML Diagramming Guide

100 Index

synchronization state 82
synchronous message 48, 72

T
template 39, 45
terminate 73
terminate node 73, 76, 83
terminate psuedostate 83
time signal 27
timing diagram 93
to interface path 56
transition 82
trees 69

U
unidirectional association 40
unnavigable association 36
unsupported UML 93
use case diagram 89

UML Diagramming Guide

 Index 101

UML Diagramming Guide

102 Index

UML Diagramming Guide