Before we start diving into more technical articles and lots of code, I wanted to make sure that all the Buzz Words of the .NET framework is understood by all readers and maybe answer couple of questions regarding where all the pieces fit.

First, we defined the CLR (Common Language Runtime) in the last article, now we will attack the CTS (Common Type System).

The Common Type System (CTS)

Probably one of the most attractive features of the .NET CLR in that it is based on the CTS, which provides a very rich and standard set of data types. The CTS is object oriented by design but also supports procedural and functional languages as well.

The CTS is what allows .NET to provide a unified programming model, and to support multiple languages. The CTS supports two general categories of types, each of which actually have a number of subcategories.

A Type System

A type describes a value and specifies a contract that all values of that type must support. Because the CTS is Object Oriented and also supports functional and procedural languages, it supports two different kinds of entities:

1. Reference Types
2. Value Types.

Value types directly contain their data, and instances of value types are either allocated on the stack or allocated inline in a structure.

Value types can be built-in (implemented by the runtime), user-defined, or enumerations.

Reference Types, store a reference to the value's memory address, and are allocated on the heap. Reference types can be self-describing types, pointer types, or interface types. The type of a reference type can be determined from values of self-describing types.

Self-describing types are further split into arrays and class types. The class types are user-defined classes, boxed value types, and delegates. (A lot more on that in later articles)

Value Types

Value types represent values allocated on the stack. They describe values that can be represented as a sequence of bits. They cannot be NULL and must always contain a value. When value types are passed into a function, they are passed by value, meaning that a copy of the value is made prior to the function execution. This also means that the value cant change, no matter what goes on in the function. Value types also include intrinsic types. Since intrinsic types are small in size and dont consume much memory, the resource cost of making a copy is negligible and outweighs the performance drawbacks of object management and garbage collection. Examples of value types include:

• Primitives
• Structures
• Enumerations.

You can also create a value type by deriving a class from System.ValueType.

Reference Types

Reference types describe values that contain references to heap-based objects and they can be null. There are four kinds of Reference Types:

• An object type is a reference type of self-describing value. Some object types (e.g. abstract classes) are only a partial description of a value.
• An interface type is always a partial description of a value, potentially supported by many object types
• A pointer type is a compile time description of a value whose representation is a machine address of a location.
• Built-in types

These types are passed by reference, meaning that an actual address of a memory location is passed when you pass one into a function, Reference types in .NET are allocated on the managed heap, which means that it is managed by the CLR and garbage-collected by the CLR.

The Common Language Specification (CLS)

The CLS is simply a specification that defines the rules to support language integration in such a way that programs written in any language, yet can interoperate with one another, taking full advantage of inheritance, polymorphism, exceptions, and other features. These rules and the specification are documented in the ECMA proposed standard document, "Partition I Architecture", available here.

The great news for Delphi and other languages is that there is essentially a level playing field, in that all CLS compliant languages can access all the features of the CLR, use the BCL (Borland Component Library, Oops! I mean Base Class Library), and be managed by the CLR.

Assemblies

Thanks to Assemblies we no longer have DLL Hell, welcome to Assembly Hell. J

An Assembly is the unit of deployment, of security, of versioning, and of scope for the types contained within it, pretty close to what a Package is in Delphi and C++Builder.

It is self-describing. It is typically one physical DLL or EXE, in the Windows PE file format, but could be made up of multiple files.

An assembly must have a manifest (See below) that describes its contents and it usually contains MSIL, resources, and metadata describing the types contained within the assembly.

An assembly is made up of four elements:

7        Manifest

7        Metadata describing the types

7        Module(s)

7        Resources

Metadata

Data that describes data.

In the .NET sense, metadata is the information that is stored in the Assembly to make it self-describing.

So, in .NET, metadata essentially describes the elements of the Common Type System that you use in your application, as well as the information that the runtime needs to do its stuff in the areas of type safety and security.

 Manifest

The Manifest is what describes the assembly itself.

The Manifest contains:

7        A simple name

7        A four-part version number of the form Major.Minor.Build.Revision

7        Publishers private key

7        Culture (Locale)

7        List of files that make up the assembly

7        List of dependent assemblies

7        Permission requests

7        Exported types

7        Resources

 Strong Names

An assembly can be given a unique name when the publisher can include a public cryptographic key.

The key goes into the Manifest. If the assembly has a public key, it is said to have a Strong Name or shared name. In addition, the publisher can digitally sign the assembly, using a public and private key. This prevents tampering with the Assembly.

Digitally signing an Assembly places a cryptographic hash of the contents of each file in the Assembly. This is verified at run-time to ensure that it has not been corrupted or tampered with.

Signing takes place with the Strong Name tool (sn.exe) included in the .NET Framework SDK. We will discuss this in more detail in a separate article with an example.

 Managed Code

Managed Code is MSIL code that is executed directly by the Runtime (the CLR's Virtual Execution System (VES)) as opposed to native X86 code.

In .NET application logic is encoded in IL when emitted from a .NET Language compiler and then "managed" by the runtime. This means that the CLR can perform services to the managed code and types such as debugging, exception handling, serialization, security and garbage collection.

Unmanaged Code

Unmanaged code is X86 code that you have been writing for ages. Like COM, COM+, C++ libraries, Win32 code, Windows API OS functionality. The term unmanaged indicates that this code cannot make use of the .NET Runtime Services.

COM is like SMOKING: If you have not started, you should not start now, if you are already doing it, it is time to stop. J

 Managed Data

 Managed data is data that is under control of the CLR's Garbage Collector. All dynamically allocated data is allocated from the managed heap and is thus, managed data. This data has metadata associated with it and thus, is self-describing. The CLR performs memory layout of the data and can also operate on it at runtime to discover type data through a process known as Reflection. (Many articles on Reflection will follow).

 Global Assembly Cache and Shared Assemblies

The GAC is a shared repository of assemblies that is machine-wide. This allows multiple applications to share an assembly. All assemblies that are to be placed in the GAC must have a Strong Name. The GAC also supports versioning. You may have multiple versions of the same assembly (each with a different version number) executing simultaneously.

If my Application ABC uses an Assembly called A.dll and another application XYZ uses the same Assembly A.dll. If a new version of ABC is released with a revised A.dll Assembly there is no need for XYZ to link with the new A.dll Assembly. Both A.dll Assemblies can co-exist in different directories and the GAC can notify the application running of which one is needed to run an app.

 Write to you later, till then, have fun

About Falafel Software Inc:

Falafel Software is all about making the most of software development technology in order to complete the project on time and on budget with best possible user experience. Falafel Software offers a comprehensive suite of software development solutions ranging from strategy to design to implementation that businesses need in order to realize high returns on their investment.

Copyright ) 2003 Alain Tadros, Falafel Software Inc.
ALL RIGHTS RESERVED. NO PART OF THIS DOCUMENT CAN BE COPIED IN ANY FORM WITHOUT THE EXPRESS, WRITTEN CONSENT OF THE AUTHOR.