Responding to rao...
How do you show multiple/consecutive actions in a sequence
diagram. For example if an action starts at a point, say showing
a form which then followed by another action say right-clicking
and creating new form.
Sequence Diagrams do not describe actions. (One uses Activity
Diagrams or Abstract Action Language to describe what goes on within
entity actions.) They describe collaborations expressed as messages
between entities. Those messages <may> trigger response
actions in the receiver. At best they only identify actions when the
message is a Call Event because the message is also the method
(action) signature. The purpose of the SD is to describe the overall
flow of control between entities by daisy-chaining messages together
in a particular sequence.
In your example the actions seem to be triggered by external user
requests (e.g., "show form" or "create form"). In that case it is
conventional to represent that external actor as a swimlane. The
messages mapping to the key clicks would then be directed to the
software entities whose actions would respond to the actor's request
(i.e., to other swimlanes representing the software entities).
Note that the reason the SD is called a Sequence Diagram is
because the messages cascade in a particular sequence as one moves
down the swimlanes. So the notion of consecutive actions is implicit
in the way one constructs the diagram (e.g., messages lower down in
the diagram are issued after messages higher up are consumed).
How should such action be displayed in UML sequence diagram.
Should I draw two separate UML diagrams but the two are related
or some how combine the UML diagram but don't know how?
That depends. Is creating a new form always done after the user has
shown a form? If so, one uses a single SD to capture that particular
sequence. OTOH, if the showing of forms and creation of forms are
completely different activities (e.g., steps in different, unrelated
use cases) then one would probably use separate diagrams because the
sequencing is random. If the requests are closely related but could
be issued in any order, then one should probably use a Collaboration
Diagram since the external interactions are inherently asynchronous.
--
Life is the only flaw in an otherwise perfect nonexistence
-- Schopenhauer
H. S. Lahman
H.la...@verizon.net
software blog: http://pathfinderpeople.blogs.com/hslahman/index.html