Chapter7

Front 1

_ is any error that, if not removed, could cause a software system tofail to meet its users’ needs.

Back 1

A software defect

Front 2

____ is the degree to which a software product meets the needs of its users.___focuses on defining, measuring, and refining the quality of the devel-opment process and the products developed during its various stages. These products—including statements of requirements, flowcharts, and user documentation—are known as__

Back 2

Software quality, Quality management, deliverables.

Front 3

___ is a set of interrelated components—including hardware,software, databases, networks, people, and procedures—that collects and processes dataand disseminates the output.

Back 3

A business information system

Front 4

Another type of business information system is the ____,which is used to improve decision making in a variety of industries. A DSS can help developaccurate forecasts of customer demand, recommend stocks and bonds for an investmentportfolio, or schedule shift workers in such a way as to minimize cost while meetingcustomer service goals. A software defect in a DSS can result in significant negativeconsequences for an organization and its customers.

Back 4

decision support system (DSS)

Front 5

For example, steel manufacturers use process control software to capture data fromsensors about the equipment that rolls steel into bars and about the furnace that heats thesteel before it is rolled. Without process control computers, workers could react to defectsonly after the fact and would have to guess at the adjustments needed to correct the process.Process control computers enable the process to be monitored for variations from operatingstandards (e.g., a low furnace temperature or incorrect levels of iron ore) and to eliminateproduct defects before they affect product quality. Any defect in this software can lead todecreased product quality, increased waste and costs, or even unsafe operating conditionsfor employees. (Some consequences of these defects are discussed later in this chapter.)

Back 5

kkkk

Front 6

The liability of manufacturers, sellers, lessors, and others for injuries caused by defec-tive products is commonly referred to as ___

Back 6

product liability.

Front 7

___means that the defendant is held responsible for injuring another person,regardless of negligence or intent. The plaintiff must prove only that the software productis defective or unreasonably dangerous and that the defect caused the injury.

Back 7

Strict liability

Front 8

____ is the failure to do what a reasonable person would do, or doing somethingthat a reasonable person would not do.

Back 8

Negligence

Front 9

___ assures buyers or lessees that a product meets certain standards of quality.A warranty of quality may be either expressly stated or implied by law. Express warrantiescan be oral, written, or inferred from the seller’s conduct

Back 9

A warranty

Front 10

If the product fails to meet the terms of its warranty, the buyer or lessee can sue for____ Of course, most dissatisfied customers will first seek a replacement,a substitute product, or a refund before filing a lawsuit.

Back 10

breach of warranty.

Front 11

___occurs when a seller or lessor either misrepresentsthe quality of a product or conceals a defect in it. For example, if a cleaning product isadvertised as safe to use in confined areas and some users subsequently pass out from theproduct’s fumes, they could sue the seller for intentional misrepresentation or fraud.

Back 11

Intentional misrepresentation

Front 12

Many software companies have adopted a ___—astandard, proven work process that enables systems analysts, programmers, project man-agers, and others to make controlled and orderly progress in developing high-qualitysoftware.

Back 12

software development methodology

Front 13

As with most things, it is usually easier and cheaper to avoid software problems at thebeginning than to attempt to fix the damages after the fact. Studies have shown that thecost to identify and remove a defect in an early stage of software development can be upto 100 times less than removing a defect in a piece of software that has been distributed tocustomers.21 If a defect is uncovered during a later stage of development, some rework ofthe deliverables produced in preceding stages will be necessary. The later the error isdetected, the greater the number of people who will be affected by the error. Thus, thegreater the costs will be to communicate and fix the error. Consider the cost to communi-cate the details of a defect, distribute and apply software fixes, and possibly retrain endusers for a software product that has been sold to hundreds or thousands of customers.Thus, most software developers try to identify and remove errors early in the developmentprocess not only as a cost-saving measure but as the most efficient way to improve softwarequality. A product containing inherent defects that harm the user may be the subject of aproduct liability suit. The use of an effective methodology can protect software manufac-turers from legal liability in two ways. First, an effective methodology reduces the numberof software errors that might occur. Second, if an organization follows widely accepteddevelopment methods, negligence on its part is harder to prove. However, even a successfuldefense against a product liability case can cost hundreds of thousands of dollars in legalfees. Thus, failure to develop software carefully and consistently can be serious in terms ofliability exposure.

Back 13

kkk

Front 14

____ refers to methods within the development cycle designed toguarantee reliable operation of the product. Ideally, these methods are applied at each stageof the development cycle. However, some software manufacturing organizations without

Back 14

Quality assurance (QA)

Front 15

Software is developed in units called subroutines or programs. These units, in turn, are com-bined to form large systems. One approach to QA is to test the code for a completed unitof software by actually entering test data and comparing the results to the expected results.This is called ____ There are two forms of dynamic testing:

Back 15

dynamic testing.

Front 16

__testing involves viewing the software unit as a device that hasexpected input and output behaviors but whose internal workings are unknown(a black box). If the unit demonstrates the expected behaviors for all the inputdata in the test suite, it passes the test. Black-box testing takes place withoutthe tester having any knowledge of the structure or nature of the actual code.For this reason, it is often done by someone other than the person who wrotethe code.

Back 16

Black-box

Front 17

_____testing treats the software unit as a device that has expected inputand output behaviors but whose internal workings, unlike the unit in black-boxtesting, are known. White-box testing involves testing all possible logic pathsthrough the software unit with thorough knowledge of its logic. The test datamust be carefully constructed so that each program statement executes at leastonce. For example, if a developer creates a program to calculate an employee’sgross pay, the tester would develop data to test cases in which the employeeworked less than 40 hours, exactly 40 hours, and more than 40 hours (to checkthe calculation of overtime pay).

Back 17

White-box

Front 18

____—Special software programs called static analyzers are runagainst the new code. Rather than reviewing input and output, the static ana-lyzer looks for suspicious patterns in programs that might indicate a defect.

____—After successful unit testing, the software units are com-bined into an integrated subsystem that undergoes rigorous testing to ensurethat the linkages among the various subsystems work successfully.

____—After successful integration testing, the various subsystemsare combined to test the entire system as a complete entity.

______—Independent testing performed by trained endusers to ensure that the system operates as they expect.

Back 18

Static testing, Integration testing, System testing, User acceptance testing

Front 19

_____—developed by the Software EngineeringInstitute at Carnegie Mellon—is a process improvement approach that defines the essentialelements of effective processes. The model is general enough to be used to evaluate and improve almost any process, and a specific application of CMMI—CMMI-Development(CMMI-DEV)—is frequently used to assess and improve software development practices.

Back 19

Capability Maturity Model Integration (CMMI)

Front 20

_____ is one whose failure may cause injury or death. The safe operationof many safety-critical systems relies on the flawless performance of software; such systemscontrol automobiles’ antilock brakes, nuclear power plant reactors, airplane navigation,roller coasters, elevators, and numerous medical devices, to name just a few.

Back 20

A safety-critical system

Front 21

The key to ensuring that these additional tasks are completed is to appoint a projectsafety engineer, who has explicit responsibility for the system’s safety. _____ uses a logging and monitoring system to track hazards from a project’s start to its finish.

Back 21

The safety engineer

Front 22

___ is the probability of an undesirableevent occurring times the magnitude of the event’s consequences if it does happen.

Back 22

Risk

Front 23

Another key element of safety-critical systems is _____ the provision of multipleinterchangeable components to perform a single function in order to cope with failures anderrors.

Back 23

redundancy,

Front 24

An example of a simple redundant system would be an automobile with a spare tireor a parachute with a backup chute attached. A more complex system used in IT is a re-dundant array of independent disks (RAID), which is commonly used in high-volume datastorage for file servers. RAID systems use many small-capacity disk drives to store largeamounts of data to provide increased reliability and redundancy. Should one of the drivesfail, it can be removed and a new one inserted in its place. Since the data has also beenstored elsewhere, data on the failed disk can be rebuilt automatically without the serverever having to be shut down. N-version programming is a form of redundancy involving the simultaneous executionof a series of program instructions by two different systems.

Back 24

kkk

Front 25

____ is the probability of a component or system performing without failure overits product life. For example, if a component has a reliability of 99.9 percent, it has onechance in one thousand of failing over its lifetime.

Back 25

Reliability

Front 26

For example, automobile manufacturers have beenknown to weigh the cost of potential lawsuits against that of a recall. Drivers and passengersin affected automobiles (and, in many cases, the courts) have not found this approach tobe ethically sound. Manufacturers of medical equipment and airplanes have had to makesimilar decisions, which can be complicated if data cannot pinpoint the cause of the prob-lem. For example, there was great controversy in 2000 over the use of Firestone tires onFord Explorers after numerous tire blowouts and Explorer rollovers caused multiple injuriesand deaths. However, it was difficult to determine if the rollovers were caused by poor auto-mobile design, faulty tires, or improperly inflated tires. Consumers’ confidence in bothmanufacturers and their products was nevertheless shaken.

Back 26

kkkk (ertiti zemot)

Front 27

The International Organization for Standardization (ISO), founded in 1947, is a worldwidefederation of national standards bodies from 161 countries. The ___issued its 9000 seriesof business management standards in 1988. These standards require organizations todevelop formal quality-management systems that focus on identifying and meeting theneeds, desires, and expectations of their customers. The ISO 9000 standard serves as a guide to quality products, services, and manage-ment.

Back 27

ISO

Front 28

___is an important technique used to developISO 9000–compliant quality systems by both evaluating reliability and determining theeffects of system and equipment failures. The goal of FMEA is to identify potential design and process failures earlyin a project, when they are relatively easy and inexpensive to correct.

Back 28

Failure mode and effects analysis (FMEA)