Solutions

These programming issues have perplexed developers for a number of years , and several techniques have already been developed to address them. For example:

• Representation standards, such as eXternal Data Representation (XDR) and Network Data Representation (NDR), address the issue of passing data types between different machines. These standards compensate for issues such as big/little endian and different word sizes.

• Language standards, such as ANSI C, allow the distribution of source code across compilers and machines. They permit source code to be recompiled on different machines and with different libraries without the need to modify the source files.

• Architecture standards, such as the Distributed Computing Environment's (DCE) Remote Procedure Call (RPC), OMG's CORBA, and Microsoft's (Distributed) Component Object Model (COM/DCOM), address the issue of calling methods across language, process, and machine boundaries.

• Execution environments, such as those provided by SmallTalk's and Java's Virtual Machines (VMs), allow code to execute on different physical machines by providing a standardized execution environment.

All of these approaches provide significant benefits to the application developer, but unfortunately none has solved ”or even attempted to address ”all the problems associated with a distributed computing environment. In particular, the issue of language interoperability deserves more attention. Language interoperability entails more than just a standardized calling model, such as those provided by COM and CORBA; it also involves a schema that allows classes and objects in one language to be used as first-class citizens in another language. Achieving this level of interoperability is a major goal of Microsoft's .NET Framework.