the process by which the analyst team ensures that every aspect of the system performs as it should.
the process consisting of those activities associated with ensuring that the new system is fully functional and operational, as well as those activities associated with turning over control of the new system to the end users.
tests in which the code is actually inspected
the process of inspecting the atual source code for the occurrence of certain types of errors commonly associated with the language in which the program has been written.
a close examination of the embedded logic in the code.
an application test in which one or more programmers, not responsible for the actual writing of the code, work through a hard copy of source code, mentally simulating the control flow.
an application test that focuses on ascertaining the successful execution of each application module prior to integrating it with other tested modules. Also referred to as the unit test.
tests in which the module is not inspected and is usually not tested by its author. The module is treated like a black-box and fed typical (and sometimes not so typical) input, while the resultant output is evaluated for conformance to expectations.
an integration strategy in which individually tested modules are assembled into low-level subsystems that are retested and then assembled into higher-level modules until the application is complete.
a driver that simulates the control environment for the module under test by providing simulated input or receiving module output.
a testing strategy in which the highest-level control module is tested first and the lower-level modules are simulated by a program stub, a two or three line module containing no logic, that is designed to simply accept control from a high-level module and return it to that module. Also referred to as stub testing.
Four conversion strategies:
an approach to system conversion in which the old system is simply turned off and the new system is turned on in its place. Also referred to as the slam-dunk or cold-turkey strategy.
a system conversion strategy in which the old and new systems are run simultaneously until the end users and project coordinators are fully satisfied that the new system is functioning correctly and the old system is no longer necessary.
a system conversion strategy in which the conversion to the new system is done, using either a direct or parallel method, at a single location first.
a system conversion strategy that attempts to take advantage of the best features of both the direct and the parallel appproaches while minimizing the risks involved; an incremental approach that allows for the new system to be brought online as a series of functional components that are logically ordered so as to minimize disruption to the end users and the flow of business. Also known as gradual conversion.
a type of parallel conversion in which a predetermined date for stopping the parallel operation is set.
a type of parallel conversion in which some predetermined method of phasing in each piece of the new system and turning off a similar piece of the old system is employed.
postimplementation maintenance phase
the SDLC phase that consists of correcting errors or faults in the system, providing changes to effect performance improvement, and adapting the system to changes in the operating or business environment.
change request maintenance
the modification of existing system functions with an eye toward identifying and implementing changes to the system that add or enhance performnace and functionality.
change control steering committee
a steering committee established for the purpose of reviewing and approving any change requests made for a particular system.