The purpose of a TALC2000 "Menu" is to allow you to group your test cases into logical groupings that correspond to the groupings in your application software (e.g. data entry, enquiry, listings etc).

You can create as many TALC2000 menus as you wish, and each menu can contain up to 90 separate test cases.

Before you begin recording your first test case, if you have not already done so, divide your test cases into appropriate groups and then create a TALC2000 menu for each group of tests.

Select as the active menu the menu you will require for the recording of your first test cases.

Test case creation is simple. The test is executed manually, using the mouse and keyboard. Each mouse or keyboard action that you perform will be recorded and you will also record your application software responses to the input you are creating by clicking on a button in TALC2000's Floating Toolbar. This single action will record information about the size and location of your test window, and the text and attributes that are displayed.

TALC2000 creates a character based image of the information displayed in the window, to correspond with what would be displayed on a "dumb" terminal.

When you have completed the test manually, the information you have recorded is saved by a very simple process which allows you to add details of the test which you have just performed to the active TALC2000 menu.

As well as naming the test with up to 40 characters of description, you can also optionally add up to 2,000 characters of additional documentation for the test, to define the test purpose, the test environment, any test prerequisites and any other information which you find useful to retain.

Each test which you record and save should be designed to test a single function. If you are testing a large application system, this will result in probably thousands of separate test cases spread across many different TALC2000 menus.

Once you have recorded your individual test cases, you will then want to create one or more TALC2000 "Scripts" which will dynamically link these test cases together for you, so that they may be played back as a single complete test. This linking process is extremely simple (little more than "point and click") and requires no programming knowledge.

• Make sure your test data base is re-set to contain just the data required to begin the test run.

• Select from a TALC2000 menu the script that you wish to run to execute the complete test.

• Click the "Play" button on the TALC2000 Floating Toolbar.

As well as beginning test playback manually by clicking on the "Play" button, you may also set up TALC2000 to begin test playback at a pre-defined time, or following some specific event which can display a message on your terminal. (For example, "End of day completed").

The tests can run unattended, overnight if desired, and the validation processing can be undertaken after the tests have completed when you next return to your PC.

Although the above test creation and execution scenario is extremely simple, about 90% of your tests will be recorded and linked together for playback in this way.

Where you cannot achieve the required level of test control using the methods outlined above, TALC2000 provides additional ways of creating tests, of controlling test execution, and of performing the necessary validation.

TALC2000 also provides many special features to help with Year 2000 testing. The most important of these is the ability to process date fields of any type as variables that are evaluated at the time the test is run.

This feature allows you to create a test case and run it across multiple different processing time horizons without the need for any test maintenance. By selection of a single TALC2000 menu entry, that same test case can be run in your Year 2000 environment, using the required new date formats for input and output.

Where required, TALC2000 can automatically generate high volumes of test transactions for you, based on definitions of the transaction that you provide, and variable data contained in PC text files.

This feature can be particularly useful when it is necessary to generate high volumes of data for testing screen or report paging, maximum number of entries in a table, list box scrolling and so on.

To provide additional control during test execution, TALC2000 supports a comprehensive "Event/Action" module. This module allows you to monitor asynchronously for events which occur during test playback. These events may be associated with the test, or completely external to it.

An "Event" is typically a string of characters appearing at some point on the screen. "Actions" associated with each event can include insertion of additional keystrokes, starting or stopping playback, suspending operations, enabling or disabling other events and many additional functions. You associate an "Event/Action" list with a particular test case by using the Event/Action entry on the TALC2000 menu.

Setting up an Event/Action list can be done at any time after the test is recorded, and does not require editing of the test. The list is simply added to the test case, and if such a list exists, TALC2000 will load and process it.

Event/Action processing allows you to check for any condition during test execution, and return a value to your executing script that identifies which event has been found. The script may then use this information to control the path of test execution. (Test execution paths can also be controlled without the use of Event/Action lists, simply by making use of the "If any errors" command in the script processor. This provides a pass/fail result for the previously executed test, and can be used for the majority of script control requirements).

As the test is executed, as well as checking in real time the pass/fail conditions of each test, the results of each test are saved so that you can review them at a convenient time.

The test validation procedure is fast, and depending on the speed of your PC and the validation techniques selected, up to about 5,000 test cases per hour can be completely checked.

The test validation process is initiated by simply clicking on a button on the TALC2000 Toolbar, and then selecting from the three available methods of verification processing. The verification process produces a summary by script of all test cases, and their pass/fail condition.

Where a given test case has failed, the information contained in that test case may be expanded down to individual validation components and inspected in a variety of ways to determine what it is that has gone wrong.

Once an error has been identified, it may be documented, prioritised and categorised as to the reason for the error.