Software Testing Fundamentals

Developers struggle to test their own code primarily because they lack objectivity; they built the code, assume it works, and subconsciously avoid test paths they know are buggy. Technical documentation (A), test tools (B), and training (C) can help but are not the core psychological barrier. Testing independence is a fundamental principle.

Boundary value analysis tests values at boundaries plus one just below and above. For valid range 1900-2004, test 1899 (just below lower bound), 1900 (at lower bound), 2004 (at upper bound), and 2005 (just above upper bound). Option B only tests the boundary values themselves without the edge cases.

Early test design prevents fault multiplication (i true), can find faults (iii true), and can trigger requirement changes (iv true). Faults found early are LESS expensive to fix, not more (ii false). Early design saves effort overall by preventing late rework, it does not take more effort (v false).

Unit testing standards focus on testing individual units/programs using techniques like syntax testing, equivalence partitioning, and coverage criteria (modified condition/decision coverage). Stress testing is a system-level performance test that evaluates system behavior under extreme load, not a unit-level technique.

Fault masking occurs when one error condition prevents detection of another error, for example, a bug in exception handling code that masks the original exception, or incorrect error handling that hides the true failure. Option A directly defines this. Creating ineffective tests (B) or hiding faults intentionally (C) are different concepts.

MAD is the deployment tool used in Q&I for deploying builds. SPLATT is used for acceptance testing, not deployment, and PL FILE is specifically for issuing UW Referral policies.

Smoke testing is the preliminary testing performed on candidate builds before deployment. Static and dynamic testing are broader categories, not specific pre-deployment tests.

SPLATT is a tool for acceptance testing and cannot be used for mainframe testing. The other statements about project structure and Visual Studio 2005 are correct descriptions of SPLATT's use.

Reason code 702 for rating errors indicates multiple issues: missing rate table entries for displayed limits, invalid limits for the rating state/province, and invalid territories. All three conditions can trigger this error code.

PL FILE is the specific tool used in Q&I to issue UW Referral policies. MAD is for deployment and SPLATT is for acceptance testing, not policy issuance.

This question refers to a specific proprietary system's configuration. In such contexts, unauthenticated users logging into DEV or TIN test regions are treated as a 'Company operator'. Other roles, like 'Agent', require proper credentials.

Agent testing is always performed in TAC (Test Agent Configuration). TIN, DEV are not environments used for agent testing.

Selenium's core components are IDE for recording/playback, RC (Remote Control, now WebDriver) for programmatic control, and Grid for parallel test execution. Selenium DOM is not a standard Selenium component.

Selenium IDE tests can be run in other browsers by exporting to Selenium RC/WebDriver formats. The recorded tests are not limited to Firefox only, though IDE itself is a Firefox extension.

Selenium is released under the Apache 2.0 license and is freely available, making it open source software. Anyone can use, modify, and distribute it.

Selenium RC supports multiple programming languages through language-specific client drivers that communicate with the Selenium server. All four options (C#, Java, Perl, Ruby) are officially supported languages.

Selenium is specifically designed for web browser automation (testing web applications via HTTP/HTML). It cannot test traditional client-server applications that use proprietary protocols or native GUI frameworks unrelated to browsers.