Availability Guide for Application Design

Designing Applications for Change
Availability Guide for Application Design—525637-004
10-22
Supporting Implementation
Supporting Implementation
Many implementation support techniques you can use make application upgrade
easier and enhance the availability of your application. Such techniques include:
•
Using extensible and variable procedures
Usng such procedures makes it easier to add functions to your procedures. Refer
to Using Extensible and Variable Procedures, following.
•
Considering Portability Requirements
Programming languages, development tools, and target platforms continue to
proliferate. Many offer attractive new features and functionality, but most are
incompatible with existing products, further complicating implementation choices. It
is easier to upgrade an application that requires porting of selected portions when
you consider the topics discussed previously under Considering Portability
Requirements on page 10-2.
•
Using procedure names that include version identifiers
It is easier to plan your upgrade if you can easily establish compatibility issues with
other procedures. Including version identifiers as part of procedure names is one
way to provide this information.
•
Avoidance of tying your code to any specific entity
It is easier to upgrade an application module online if that module is not tied to
specific entities such as a processor or a process name. Tying a process to a
specific name, for example, can make online upgrade difficult because you must
delete the old process running in that name before you can create the new one.
Refer to Avoiding Dependencies on a Specific Entity on page 10-23.
Using Extensible and Variable Procedures
You can make it easier to add function to your application by designing procedures or
functions to be extensible or variable. In this way, you can allow for parameters that will
support future functions of the application.
Traditional design—without extensible or variable procedures—defines a procedure or
function to take a specific number of known parameters. If the procedure is defined as
accepting three parameters, you must declare it with three parameters and call it with
three parameters. Adding a parameter to the procedure means not only redefining the
procedure itself but also recoding and recompiling all procedures that call it. In other
words, much of the application must be brought down to implement the new function.
Procedures can be defined with the variable attribute in order to allow parameters to
be optional. This way, you can reserve parameters for future use. When the new
capability is added to the procedure or function, existing callers do not need to change;
they simply do not use the new capability. Access to the new capability can then be
made by coding new modules.