Guidelines and Drawing Conventions

• Pre- and post-conditions should be documented in condition marks on respective lifelines or in respective tags.
• If possible, arrange lifelines such that the message exchange occurs in a "general" left-to-right flow from the top of the sequence down to the bottom. In other words, arrange the order of lifelines to minimize message zigzagging.
• For documentation reasons the print-out of a scenario should be captured on one page.
• Divide complicated scenarios into manageable, well-documented, logically related groups of messages using partition lines.
• Extract reused portions of scenarios into separate sequence diagrams that are included using interaction occurrences.
• Limit the usage of interaction operators to a maximum of two layers.
• All message lines should be horizontal, rather than diagonal. Asynchronous messages between blocks have an open arrowhead and synchronous, reflexive messages have a filled arrowhead.
• Stereotype messages according to their associated protocol (e.g. M1553, Ethernet, etc …).
• Use the statechart diagram action language to express constraints to provide the best transition to statechart diagrams.
• For simplicity, restrict the use of condition marks to represent state changes in reaction to State/Mode Change Operational Contract messages. If used for this purpose, condition marks should match the names of states in statechart diagrams.
• Do not show operations on the actor lifelines.

Naming Conventions

• The diagram shall have the associated use case name in plain text at the top of the diagram.
• Naming convention for sequence diagrams: SD_Ucd<Nr>_Uc<Nr>_<SequenceName>
• The names of reflexive (loop back) messages (= operations) do not contain a special prefix, though they must begin with a lower-case letter.