ASD Current Curriculum V2 (2006-2007)
The following topic list constitutes the Principles of Symbian OS curriculum. A full set of supporting course materials, broken down into topic-based modules, is available for universities. Please contact Symbian Academy for further details.
The questions contained within the Accredited Symbian Developer examination are derived from the Principles of Symbian OS curriculum. Any course adhering to the curriculum will therefore prepare a student to become an Accredited Symbian Developer.
- C++ Language Fundamentals
- Classes And Objects
- Class Design And Inheritance
- Symbian OS Types & Declarations
- Cleanup Stack
- Object Construction
- Descriptors
- Dynamic Arrays
- Active Objects
- System Structure
- Client Server
- File Server, Store & Streams
- Sockets
- Tool Chain
- Platform Security
- Binary Compatability
1. C++ Language Fundamentals
Evaluates the basic understanding of writing a C++ program, the language structure, syntax and semantics.
Summary
The following objectives cover:
• built in types
• statements and expressions
• function use and syntax
• dynamic memory
• basic tools (compiler and linker)
Objectives
Types
1.1. Understand that C++ has a number of different types in various categories e.g. integral, arithmetic
1.2. Recognize typedef as defining a synonym for an existing type but not a new type in itself
1.3. Recognize enumeration types as user-defined value sets
1.4. Specify the advantages of const over #define
1.5. Understand the use and properties of the C++ reference type
1.6. Specify the difference between pointers and references
1.7. Understand the semantics of pointer arithmetic
1.8. Recognize pointer operations and the purpose of the NULL pointer value
1.9. Differentiate const pointers and pointers to const
Statements
1.10. Know the use and properties of the declaration and definition statements
1.11. Recognize the use of the extern keyword
1.12. Cite and understand initialization of variables and their scope
1.13. Understand the purpose, syntax and behavior of C++ loop statements (while, for and do)
1.14. Specify the behavior and effect of the continue and break keywords in a loop
1.15. Specify the syntax and behavior of C++ conditional statements (if and switch)
Expressions and Operators
1.16. Specify the syntax and meaning of unary and binary operator expressions
1.17. Understand the difference between precedence and associativity
1.18. Recognize common operator categories including logical, prefix and postfix
1.19. Understand the operator precedence and associativity rules
Functions
1.20. Understand the syntax of a function prototype
1.21. Cite the purpose of the inline keyword
1.22. Understand the rules for passing default arguments and an unspecified number of arguments to a
function
1.23. Recognize value, reference, array and pointer parameter passing
2. Classes And Objects
Measures the understanding of C++ object-oriented programming support and class semantics
Summary
The following objectives cover:
• scope rules, data abstraction and structures
• construction and destruction
• access control
Objectives
Scope and C++ Object-Oriented Programming (OOP) Support
2.1. Understand the scope and lifetime properties of blocks and namespaces
2.2. Understand C++ support for data abstraction
2.3. Specify the attributes of a C++ object with respect to object-oriented programming
2.4. Know the syntax of a class declaration
2.5. Cite the differences between a class and an object
2.6. Differentiate between basic data structures (struct keyword) and classes
Constructors and Destructors
2.7. Understand the order of construction and destruction for the class and its member variables
2.8. Recognize implicit constructor invocation (overloading and pattern matching) and the role the explicit keyword plays in constructor declaration
2.9. Specify what the compiler automatically generates for user-defined classes
2.10. Understand the purpose and use of copy constructors (including parameter passing)
2.11. Understand the difference between assignment and initialization
2.12. Specify the required function members needed in a class to support safe member pointer data
ownership
Class Members
2.13. Describe private, protected and public access control for class members
2.14. Declare and specify the syntax for pointers to class members
2.15. Identify nested classes and their scope and life-span
2.16. Cite the scope access rules and lifetime of a nested class
2.17. Understand the scope rules and syntax of friend functions
2.18. Understand the semantics of member function and data addressing
2.19. Understand the purpose of the scope-resolution operator ::
2.20. Understand the role of the this pointer
2.21. Specify the properties of static class members
3. Class Design And Inheritance
Tests the understanding of more advanced C++ properties including design considerations when using inheritance and both dynamic and static polymorphism.
Summary
The following objectives cover:
• class relationships
• the fundamental properties of the base-derived inheritance hierarchy
• virtual methods
• templates
Objectives
Class Relationships
3.1. Understand the key benefits and purpose of inheritance
3.2. Specify the differences between composition, aggregation and inheritance
3.3. Cite the object-oriented relationship for inheritance
Inheritance
3.4. Be able to define public inheritance
3.5. Understand the scope resolution operator syntax for accessing the base-derived class hierarchy
3.6. Given a base-derived hierarchy, specify the access rules (including those of friend classes)
3.7. Describe the scope access rules and purpose of public, protected and private inheritance
3.8. Specify the implicit invocation order of constructors and destructors in a base-derived hierarchy
Dynamic Polymorphism – Virtual Methods
3.9. Specify the mechanisms of OO reuse available in C++
3.10. Be to able to state C++ support for polymorphism
3.11. Understand the purpose of and difference between overriding and overloading
3.12. Understand the use of overriding to modify behavior in base-derived class inheritance
3.13. Understand the rules and pattern-matching criteria for correct overloaded-function invocation
3.14. Identify the typical uses and behavior for operator overloading
3.15. Describe the purpose of the virtual table, citing constraints and overheads
3.16. Specify the use of virtual functions and their implementation tradeoffs
3.17. Understand how an abstract base class is implemented in C++
3.18. State the differences between interface and implementation inheritance
3.19. Understand and recognize the problems associated with multiple inheritance
3.20. Cite the implementation requirements to support the static_cast operator in user-defined
classes
Static Polymorphism and Templates
3.21. Specify the syntax for a simple function template specialization
3.22. Be able to cite the advantages of function templates (e.g. over macros)
3.23. Understand the inheritance rules and syntax supported by class templates
3.24. Understand the syntax and semantics of a template type/class declaration
4. Symbian OS Types & Declarations
Ensures that the candidate understands the fundamental Symbian OS types, naming conventions and coding style, and the usage paradigms of each basic type of Symbian OS class.
Summary
The following objectives cover the Symbian OS class naming conventions, usage and behaviour.
Objectives
The Fundamental Symbian OS Types
4.1. Know how the fundamental Symbian OS types relate to native built-in C++ types
4.2. Understand that the fundamental types should always be used in preference to the native built-in
C++ types (bool, int, float etc) because they are compiler-independent
T Classes
4.3. Know the purpose of a T class, what types of member data it may and may not own, and that it must
never have a destructor
4.4. Know what types of function a T class may have
4.5. Understand that a T class may be created on the heap or stack
4.6. Understand that a T class may be used as an alternative to the traditional C/C++ struct
4.7. Know that the T prefix is also used to define an enum.
C Classes
4.8. Recognize that a C class always derives from CBase
4.9. Know the purpose of a C class, and what types of data it may own
4.10. Understand that a C class must always be instantiated on the heap
4.11. Know that a C class uses two-phase construction and has its member data zero-filled when it is allocated on the heap
4.12. Understand the destruction of C classes via the virtual destructor defined in CBase
R Classes
4.13. Know the purpose of an R class, to own a resource
4.14. Understand that an R class can be instantiated on the heap or stack
4.15. Understand the separate construction and initialization of R classes
4.16. Understand the separate cleanup and destruction of R classes, and the consequences of forgetting
to call the Close() or Reset() method before destruction
M Classes
4.17. Know the purpose of an M class, to define an interface
4.18. Understand the use of M classes for multiple inheritance, and the order in which to derive an implementation class from C and M classes
4.19. Know that an M class should never contain member data and does not have constructors
4.20. Know what types of function an M class may include, and the unusual circumstances where it is
appropriate to define their implementation
4.21. Understand that an M class cannot be instantiated
Static Classes
5. Cleanup Stack
Demonstrates practical knowledge of the Cleanup Stack and the difference between standard C++ and Symbian OS in handling leaks and exceptions.
Summary
These objectives cover the following idioms:
• Symbian OS leaves vs standard C++ exception support
• Cleanup stack
• TRAP handlers
in terms of their purpose, the relationship between them, their usage and common pitfalls and practical programming techniques (e.g. debug macros to detect memory leaks).
Objectives
Leaves: Lightweight Exceptions for Symbian OS
5.1. Know that, pre v9, Symbian OS does not support standard C++ exceptions (try/catch/throw) but
uses a lightweight alternative: TRAP and leave
5.2. Know that leaves are a fundamental part of Symbian error handling and used throughout the system
5.3. Understand the similarity between leaves and the setjmp/longjmp declarations in C
5.4. Recognize the typical system functions which may cause a leave, including the user::LeaveXXX() functions and new(Eleave)
5.5. Be able to list typical circumstances which cause a leave (for example, insufficient memory for a
heap allocation)
5.6. Understand that new (Eleave) guarantees that the pointer return value will always be valid if a
leave has not occurred
How to Work with Leaves
5.7. Know that leaves are indicated by use of a trailing L suffix on functions containing code which may
leave (for example, InitializeL())
5.8. Be able to spot functions which are not leave-safe and those which are
5.9. Understand that leaves are used for error handling; code should very rarely both return an error and
be able to leave
5.10. Understand the reason why a leave should not occur in a constructor or destructor
Comparing Leaves and Panics
5.11. Understand the difference between a leave and a panic
5.12. Recognize that panics come about through assertion failures, which should be used to flag programming errors during development
5.13. Recognize that a leave should not be used to direct normal code logic
What Is a TRAP?
5.14. Recognize the characteristics of a TRAP handler
5.15. Understand that, for efficiency, use of TRAPs should be kept to a minimum
The Cleanup Stack
5.16. Know how to use the cleanup stack to make code leave-safe, so memory is not leaked in the event
of a leave
6. Object Construction
Evaluates knowledge of 2-phase construction in Symbian OS, safely creating objects and avoiding memory leaks.
Summary
The following objectives test an understanding of the importance, implementation and reasons for 2 phase construction, and how to ensure object destruction is efficient and safe.
Objectives
Two-Phase Construction
6.1. Know the reasons why code should not leave inside a constructor
6.2. Recognize that two-phase construction is used to avoid the accidental creation of objects with undefined state
6.3. Understand that constructors and second-phase ConstructL() methods are given private or protected access specifiers in classes which use two-phase construction, to prevent their inadvertent
use
6.4. Understand how to implement two-phase construction, and how to construct an object which derives
from a base class which also uses a two-phase method of initialization
6.5. Know the Symbian OS types (C classes) which typically use two-phase construction
Object Destruction
6.6. Know that it is not efficient or necessary to set a pointer to NULL after deleting it in destructor code
6.7. Understand that a destructor must check before dereferencing a pointer in case it is NULL, but need
not check if simply calling delete on that pointer
7. Descriptors
Test the understanding of the motivation for using descriptors, how to use buffer and pointer descriptors and when to use package descriptor classes.
Summary
The following objectives cover the characteristics of the Symbian OS descriptor classes including memory management, inheritance hierarchy, use as function parameters, conversion methods and key aspects of the API.
Objectives
Features of Symbian OS Descriptors
7.1. Understand that Symbian OS descriptors may contain text or binary data
7.2. Know that descriptors may be narrow (8-bit), wide (16-bit) or neutral (which is 16-bit since Symbian
OS is built for Unicode)
7.3. Understand that descriptors do not dynamically extend the data they reference, so will panic if too
small to store data resulting from a method call
The Symbian OS Descriptor Classes
7.4. Know the characteristics of the TDesC, TDes, TBufC, TBuf, TPtrC, TPtr, RBuf and HBufC descriptor classes
7.5. Understand that the descriptor base classes TDesC and TDes implement all generic descriptor manipulation code, while the derived descriptor classes merely add construction and assignment code
specific to their type
7.6. Identify the correct and incorrect use of modifier methods in the TDesC and TDes classes
7.7. Recognize that there is no HBuf class, but that RBuf can be used instead as a modifiable dynamically allocated descriptor
The Inheritance Hierarchy of the Descriptor Classes
7.8. Know the inheritance hierarchy of the descriptor classes
7.9. Understand the memory efficiency of the descriptor class inheritance model and its implications
Using the Descriptor APIs
7.10. Understand that the descriptor base classes TDesC and TDes cannot be instantiated
7.11. Understand the difference between Size(), Length() and MaxLength() descriptor methods
7.12. Understand the difference between Copy() and Set() descriptor methods and how to use assignment correctly
Descriptors as Function Parameters
7.13. Understand that the correct way to specify a descriptor as a function parameter is to use a reference, for both constant data and data that may be modified by the function in question.
Correct Use of the Dynamic Descriptor Classes
7.14. Identify the correct techniques and methods to instantiate an HBufC heap buffer object
7.15. Recognize and demonstrate knowledge of how to use the new descriptor class RBuf
Common Inefficiencies in Descriptor Usage
7.16. Know that TFileName objects should not be used indiscriminately, because of the stack space
each consumes
8. Dynamic Arrays
Measures proficiency in the use of Symbian OS dynamic arrays in preference to standard C++ arrays, and the choice of dynamic array class depending on desired usage and characteristics of array elements.
Summary
The following objectives cover an understanding of the Symbian OS dynamic array classes including their practical application, memory layout and expansion strategy (granularity).
Objectives
Dynamic Arrays in Symbian OS
8.1. Demonstrate an understanding of the basics of Symbian OS dynamic arrays (CArrayX and RArray families)
8.2. Understand the different types of Symbian OS dynamic arrays with respect to memory arrangement
(flat or segmented), object storage (within array or elsewhere), object length (fixed or variable) and
object ownership.
8.3. Recognize the appropriate circumstances for using a segmented-buffer array class rather than a flat
array class
RArray, RPointerArray or CArrayX?
8.4. Know the reasons for preferring RArrayX to CArrayX, and the exceptional cases where CArrayX classes are a better choice
Array Granularities
8.5. Understanding the meaning of array granularity
8.6. Know how to choose the granularity of an array as appropriate to its perceived use
Array Sorting and Searching
8.7. Demonstrate an understanding of how to sort and seek in dynamic arrays
8.8. Recognize that RArray, RPointerArray and the CArrayX family can all be sorted, although the CArrayX classes are not as efficient
TFixedArray
8.9. Recognize that the TFixedArray class should be preferred over a C++ array, since it gives the
benefit of bounds checking (debug-only or debug and release).
9. Active Objects
Tests understanding of why Active Objects are preferred over Threads and under what conditions, the use and implementation of simple Active Objects and their relation to the Active Scheduler.
Summary
The following objectives cover the use of active objects for non-preemptive event handling on Symbian OS including: how to derive an active object class (priority assignment and virtual overrides, and their characteristics),“conditions of use”, common errors (bad practice) and the role of the active scheduler.
Objectives
Event-Driven Multitasking on Symbian OS
9.1. Demonstrate an understanding of the difference between synchronous and asynchronous requests,
and differentiate between typical examples of each
9.2. Recognize the typical use of active objects to allow asynchronous tasks to be requested without
blocking a thread
9.3. Understand the difference between multi-tasking using multiple threads and multiple active objects,
and why the latter is preferred in Symbian OS code
Class CActive
9.4. Understand the significance of an active object’s priority level
9.5. Recognize that the active object event handler method (RunL()) is non-preemptive
9.6. Know the inheritance characteristics of active objects, and the functions they are required to implement and override
9.7. Know how to correctly construct, use and destroy an active object
The Active Scheduler
9.8. Understand the role and characteristics of the active scheduler
9.9. Know that CActiveScheduler::Start() should only be called after at least one active object
has an outstanding request
9.10. Recognize that a typical reason for a thread to fail to handle events may be that the active scheduler has not been started or has been stopped prematurely
9.11. Understand that CActiveScheduler may be sub-classed, and the reasons for creating a derived
active scheduler class
Canceling an Outstanding Request
9.12. Understand the different paths in code that the active object uses when an asynchronous request
completes normally, and as the result of a call to Cancel()
Background Tasks
9.13. Understand how to use an active object to carry out a long-running (or background) task
9.14. Demonstrate an understanding of how self-completion is implemented
Common Problems
9.15. Know some of the possible causes of a stray signal panic
9.16. Recognize that a call to User::After() blocks a thread until the time specified as a parameter
has elapsed
9.17. Recognize that a typical reason for a blocked thread may be the inappropriate use of
User::WaitForRequest()
10. System Structure
Tests knowledge of the underlying structure of the Symbian OS, including threads, processes, DLLs and memory management. Also assesses the candidate’s understanding of common system components.
Summary
The following objectives cover
• writable static data
• UIDs
• threads and processes
• memory management idioms
• system architecture including non-functional requirements
• IPC (preferred mechanisms on Symbian OS)
• recognizers
• synchronization primitives
• defensive programming
• debugging techniques
• publish and subscribe
Objectives
DLLs in Symbian OS
10.1. Know and understand the characteristics of polymorphic interface and shared library (static) DLLs
10.2. Know that UID2 values are used to distinguish between static and polymorphic DLLs, and between
plug-in types
10.3. For a shared library, understand which functions must be exported if other binary components are
to be able to access them
10.4. Know that Symbian OS does not allow library lookup by name but only by ordinal
Writable Static Data
10.5. Recognize that writable static data is not allowed in DLLs on EKA1 and discouraged on EKA2
10.6. Know the basic porting strategies for removing writable static data from DLLs
Executables in ROM and RAM
10.7. Recognize the correctness of basic statements about Symbian OS execution of DLLs and EXEs in
ROM and RAM
Threads and Processes
10.8. Recognize the correctness of basic statements about threads and processes on Symbian OS
10.9. Recognize the role and the characteristics of the synchronization primitives RMutex, RCriticalSection and RSemaphore
Inter-Process Communication (IPC)
10.10. Recognize the preferred mechanisms for IPC on Symbian OS (client–server, publish and subscribe
and message queues), and demonstrate awareness of which mechanism is most appropriate for
given scenarios
11. Client Server
Recognizes the use cases of the Client / Server model within a handset, system components utilizing the model and applies knowledge to simple Server implementations.
Summary
The following objectives cover the following:
• the role of client-server on Symbian OS
• reasons for client-server
• characteristics of the communication protocol
• runtime impact and memory overheads
• threading model
• classes used by the client-server framework
• server-side objects which must be created on start up (cleanup stack, active scheduler)
• client-server data transfer mechanism, marshalling and data lifetime requirements
• client subsessions
• system and transient servers
Objectives
The Client–Server Pattern
11.1. Know the structure and benefits of the client–server framework
11.2. Understand the different roles of system and transient servers, and match the appropriate server
type to examples of server applications
Fundamentals of the Symbian OS Client–Server Framework
11.3. Know the fundamentals of the Symbian OS client–server implementation
Symbian OS Client–Server Classes
11.4. Know the following classes used by the Symbian OS client–server framework, and basic information about the role of each:
11.5. Recognize the objects that a server must instantiate when it starts up
11.6. Understand the mechanism used to prevent the spoofing of servers in Symbian OS
Client–Server Data Transfer
11.7. Know the basics of how clients and servers transfer data for synchronous and asynchronous requests
11.8. Recognize the correct code to transfer data from a client derived from RSessionBase to a Symbian OS server
11.9. Know how to submit both synchronous and asynchronous client–server requests
11.10. Know how to convert basic and custom data types into the appropriate payload which can be
passed to the server, as both read-only and read/write request arguments
Impact of the Client–Server Framework
11.11. Understand the potential impact on run-time speed from using a client–server session and differen-tiate between circumstances where it is useful or necessary and where it is inefficient
11.12. Recognize scenarios where an implementation which uses client subsessions with the server would be recommended
11.13. Understand the impact of the context switch required when making a client–server request, and the best way to manage communication between a client and server to maximize run-time efficiency
12. File Server, Store & Streams
Identifies an understanding of the use of files, stores and streams for storing persistent and temporary data. Also measures whether the candidate has knowledge of the conditions and intended usage which determine the different classes to use for storing data.
Summary
The following objectives cover:
• RFile API methods
• RFs API methods
• Stream operators
• Stream and store classes and their use to manipulate large documents
• Difference between use of RFile and stream classes and reasons for preferring each
Objectives
The Symbian OS File System
12.1. Understand the role of the file server in the system
12.2. Know the basic functionality offered by class RFs
12.3. Recognize code which correctly opens a fileserver session (RFs) and a file subsession (RFile) and reads from and writes to the file
12.4. Know the characteristics of the four RFile API methods which open a file
12.5. Understand how TParse can be used to manipulate and query file names
Streams and Stores
12.6. Know the reasons why use of the stream APIs may be preferred over use of RFile
12.7. Understand how to use the stream and store classes to manage large documents most efficiently
12.8. Be able to recognize the Symbian OS store and stream classes and know the basic characteristics
of each (e.g. base class, memory storage, persistence, modification, etc.)
12.9. Understand how to use ExternalizeL() and operator << with RWriteStream to write an object
to a stream, and InternalizeL() and operator >> with RReadStream to read it back
12.10. Recognize that operators >> and << can leave
13. Sockets
Evaluates the ability to add communication services to an application and handling asynchronous communication events with the socket server architecture.
Summary
The following objectives cover practical usage of the RSocket API and RHostResolver class.
Objectives
Introducing Sockets
13.1. Recognize correct high-level statements which define and describe a network socket
13.2. Recognize correct statements about transport independence
13.3. Know the difference between connected and connectionless sockets
13.4. Differentiate between streamed and datagram communication and their relationship with con-
nected/connectionless sockets
The Symbian OS Sockets Architecture
13.5. Demonstrate a basic understanding of the use of sockets on Symbian OS
13.6. Recognize the characteristics of the RSocketServ, RSocket and RHostResolver classes
13.7. Understand the role and purpose of PRT protocol modules on Symbian OS
Using Symbian OS Sockets
13.8. Recognize correct patterns for opening and configuring connected and connectionless sockets
13.9. Know which RSocket API methods should be used for connected and unconnected sockets to
send and receive data
13.10. Know the characteristics of the synchronous and asynchronous methods for closing a RSocket subsession
14. Tool Chain
Measures that the candidate has a practical working knowledge of the Symbian OS tool chain and development environment.
Summary
The following objectives cover an understanding of the mmp file syntax, use of bldmake, the role of resource files, emulator settings, the difference between running code on emulator and hardware, the ABI binary standard and tool specific errors (not programming errors).
Objectives
Build Tools
14.1. Understand the basic use of bldmake, bld.inf and abld.bat
14.2. Understand the purpose and typical syntax of project definition (MMP) files
14.3. Understand the role of Symbian OS resource and text localization files
Hardware Builds
14.4. Understand that the ARM C++ EABI is an industry standard optimized for embedded application
development
14.5. Recognize basic information about the RVCT and GCCE compilers, which can be used for target
hardware builds
14.6. Understand that ARMv5 supports both 32-bit ARM and 16-bit THUMB instructions, and appreciate
the difference with respect to speed and size
Installing an Application to Phone Hardware
14.7. Recognize the package file format used for creation of SIS installation files
The Symbian OS Emulator
14.8. Understand the purpose of the Symbian OS emulator for Windows
14.9. Recognize differences between running code on the emulator and on target hardware
15. Platform Security
Assesses the understanding of the three core concepts of Platform Security: The Trust Model, Capabilities and Data Caging. Evaluates the candidate’s practical knowledge of designing, developing and distributing software
Summary
The objectives cover the following:
• Trusted Computer Base and Environment
• Capabilities and rules for dll loading and static linking
• Data caging for processes and dlls
• The capabilities required to access the protected filesystem areas
• Secure IDs and UIDs
• A practical understanding of writing a secure application
• The role of the software installer and supporting tools for sis file generation
• The signing process
• Platform security and Symbian OS servers
Objectives
The Trust Model
15.1. Understand what is meant by the axiom “a process is a unit of trust” and how Symbian OS enforces
this
15.2. Understand the purpose of the Trusted Computing Base and why it is important
15.3. Recognize that a number of Symbian OS APIs do not require security checks before they can be
used
15.4. Know that self-signed software that does not use sensitive system services is “untrusted” and can
be installed and run on the phone, although it is effectively “sandboxed”
Capability Model
15.5. Understand the relationship between capabilities and the Trusted Computing Base (TCB)
15.6. Understand the concept of user capabilities and their relationship to the Trusted Computing Environment (TCE)
15.7. Understand the relationship between the TCB/TCE, capability assignment, software install as the “gatekeeper” and the role of application signing
15.8. Recognize the different groups of capabilities, demonstrating a broad understanding of the privileges granted
15.9. Recognize how to specify platform security capabilities within an MMP file
15.10. Demonstrate an understanding of the capability rules
Data Caging
15.11. Understand how data caging works to protect all types of files via the three special directories
(\sys, \resource and \private); in particular, that data caging is used to partition all executables in the file system so, once trusted, they are protected from modification
15.12. Understand the implications of data caging for naming executable code
15.13. Recognize that data caging can be used to provide a secure area for an application’s data
16. Binary Compatability
Assesses understanding of binary compatibility and the programming and design techniques which are required to maintain binary compatibility in code modules and APIs.
Summary
The following objectives cover:
• the definition of compatibility at different levels (binary, link, source, forwards and backwards)
• what can’t be changed without breaking compatibility
• what can be changed without breaking compatibility
• best practice - designing to ensure compatibility is retained
Objectives
Levels of Compatibility
16.1. Demonstrate an understanding of source, binary, library, semantic and forward/backward compatibility
Preventing Compatibility Breaks – What Cannot Be Changed?
16.2. Recognize which attributes of a class are necessary for a change in the size of the class data not to
break compatibility
16.3. Understand which class-level changes will break source compatibility
16.4. Understand which class-level changes will break binary compatibility
16.5. Understand which library-level changes will break binary compatibility
16.6. Understand which function-level changes will break binary and source compatibility
16.7. Differentiate between derivable and non-derivable C++ classes in terms of what cannot be changed
without breaking binary compatibility
What Can Be Changed Without Breaking Compatibility?
16.8. Understand which class-level changes will not break source compatibility
16.9. Understand which class-level changes will not break binary compatibility
16.10. Understand which library-level changes will not break binary compatibility
16.11. Understand which function-level changes will not break binary and source compatibility
16.12. Differentiate between derivable and non-derivable C++ classes in terms of what can be changed
without breaking binary compatibility
Best Practice – Designing to Ensure Future Compatibility
16.13. Recognize best practice for maintaining source and binary compatibility
16.14. Recognize the coupling arising from the use of inline functions and differentiate between cases
where it will make maintaining binary compatibility more difficult and where it will be less significant
(PDF 112kb) from the ASD Primer book, courtesy of 
