Browse Prior Art Database

Comparative Software Application Testing or Bug for Bug Verification.

IP.com Disclosure Number: IPCOM000171926D
Original Publication Date: 2008-Jun-20
Included in the Prior Art Database: 2008-Jun-20
Document File: 1 page(s) / 26K

Publishing Venue

IBM

Abstract

Usually testing software involves establishing a strategy, writing test cases, executing them. Test cases vary parameters, some generate random inputs. The random input test cases are especially interesting because they can discover defects that test cases with predetermined parameters would not be able to find. If the only aspect that we are interested in, is to ensure that the old functionality still works in the new release (or broken in the both releases), we can compare test results of the old release with the new release. To put it in other words, perform a bug for bug comparison.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 1 of 1

Comparative Software Application Testing or Bug for Bug Verification .

The invention is the comparative testing that includes:

Synchronized execution of test case on two or more computers

Real-time verification

Let's say A is the previous release of software, B is the new one. During synchronized execution of the same test case against A and B, the following is observed:

A produces result X

B produces result Y

If X = Y, we will say that A and B are bug for bug compatible.

By running two software applications at the same time and verifying the test results in real time, we can construct random test cases and compare responses (or actual results) of the two or more systems. If we know that one of the systems (the template) is working correctly, the second system should be performing as well as the first one.

The following can be compared at real-time:

Final statistics


Log files, error logs
Data

There is no dependency on a predetermined result. I.e. we don't need to keep track of random inputs, a wide range of random test cases can be produced. Another way to describe the invention is by analogy with a pantograph that is used to compare two drawings. In this case, the pointer will "point" at the old release of software, the pencil at the new release. While we trace the fist drawing with a pointer, if the pencil "draws" outside of the second drawing, the comparison fails.

1