SSVV 01b Inspection

8/18/2019 SSVV 01b Inspection

1/108

OutlineSoftware quality assessment

Human (testing) methodsQuestions

Software Systems Verification and ValidationLecture 01b - Inspection

Lect. dr. Andreea Vescan

Babeş-Bolyai UniversityCluj-Napoca

2014-2015

Lect. dr. Andreea Vescan Software Systems Verification and Validation

http://find/


2/108



1 Software quality assessmentQuality assessmentExtended V model

2 Human (testing) methods

Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsPair-ProgrammingDesk check

3 QuestionsQuestions


http://find/


3/108



Quality assessmentExtended V model

Software quality

Software quality: Conformance to explicitly stated functionaland performance requirements, explicitly documenteddevelopment standards, and implicit characteristics that are

expected of all professionally developed software. [Pressman]


http://find/


4/108




Software quality


expected of all professionally developed software. [Pressman]Activities for software quality assessment


http://find/


5/108




Software quality



Static analysis


http://find/


6/108




Software quality



Static analysis

Dynamic analysis.


http://find/


7/108




Software quality



Static analysis

Based on the examination of a number of documents. It does

not involved actual execution of the code under development.

Dynamic analysis.


http://find/


8/108




Software quality



Static analysis



Includes: code review, inspection, algorithm analysis, proof of correctness.

Dynamic analysis.


http://find/


9/108




Software quality



Static analysis




Dynamic analysis.

Involved actual program execution in order to expose possible

program failures.


http://find/http://goback/


10/108




Software quality



Static analysis




Dynamic analysis.

Involved actual program execution in order to expose possible

program failures.

Includes: testing.


http://find/


11/108




Extended/Modified V model


http://find/


12/108






http://find/


13/108






Human (testing) methods

http://find/


14/108



( )Human (testing) methodsInspectionWalkthroughsPair-ProgrammingDesk check



Human (testing) methods( )

http://find/


15/108



Human (testing) methodsInspectionWalkthroughsPair-ProgrammingDesk check



O liHuman (testing) methodsH ( i ) h d

http://find/


16/108





Prevent errors - Introduction of a

verification step at the end of eachprocess.

Requirements

Objectives

External Specification

System Design

Program Structure Design

Module Interface Specifications

Code

End User

Verification

Verification

Verification

Verification

Verification

Verification

Verification


O tliHuman (testing) methodsH (t ti ) th d



17/108





Is it useful?


OutlineHuman (testing) methodsHuman (testing) methods

http://find/


18/108





Is it useful? they contribute to productivity and reliability:



http://find/


19/108





Is it useful? they contribute to productivity and reliability:The earlier errors are found, the lower the cost of correctingthe errors.



http://find/


20/108





Is it useful? they contribute to productivity and reliability:The earlier errors are found, the lower the cost of correctingthe errors.Psychological change of programmers when computer-based

testing commences.



http://find/


21/108






testing commences. more errors



http://find/


22/108

Software quality assessmentHuman (testing) methods

Questions

( g)InspectionWalkthroughsPair-ProgrammingDesk check



testing commences. more errorsHuman testing methods are:InspectionsWalkthroughsPair-programing



http://find/


23/108


Questions

( g)InspectionWalkthroughsPair-ProgrammingDesk check



testing commences. more errorsHuman testing methods are:InspectionsWalkthroughsPair-programingDesk check



http://find/


24/108


Questions

InspectionWalkthroughsPair-ProgrammingDesk check




Objective - to find errors but not to find solutions to theerrors.


OutlineS f li

Human (testing) methodsHuman (testing) methodsI i

http://find/


25/108


Questions





Objective - to find errors but not to find solutions to theerrors.Advantage - when an error is found it is usually located.


OutlineS ft lit t

Human (testing) methodsHuman (testing) methodsI ti

http://find/


26/108


Questions





Objective - to find errors but not to find solutions to theerrors.Advantage - when an error is found it is usually located.Find from 30% to 70% of the logic-design/coding errors in

programs. Lect. dr. Andreea Vescan Software Systems Verification and Validation

OutlineSoft a e alit assess e t

Human (testing) methodsHuman (testing) methodsI s ectio

http://find/


27/108


Questions





Objective - to find errors but not to find solutions to theerrors.Advantage - when an error is found it is usually located.Find from 30% to 70% of the logic-design/coding errors in

programs.Ins ection and com uter-based testin are com lementar .Lect. dr. Andreea Vescan Software Systems Verification and Validation


Human (testing) methodsHuman (testing) methodsInspection

http://find/


28/108


Questions


Inspection

Inspection - process of trying to find defects in developmentdocuments during various phases of the software developmentprocess.




http://find/


29/108


Questions


Inspection


Fagan Inspection team ([4 members])




http://find/


30/108


Questions


Inspection


Fagan Inspection team ([4 members])Moderator - duties




http://find/


31/108

So t a e qua ty assess e tHuman (testing) methods

Questions

spect oWalkthroughsPair-ProgrammingDesk check

Inspection



Distributing materials for, and scheduling the inspection

session.




http://find/


32/108

q yHuman (testing) methods

Questions

pWalkthroughsPair-ProgrammingDesk check

Inspection




session.

Leading the session




http://find/


33/108


WalkthroughsPair-ProgrammingDesk check

Inspection




session.

Leading the session

Ensuring that the errors are subsequently corrected.



( )


http://find/


34/108



Inspection




session.

Leading the session


Author of the product (analyst, designer, programmer)



H ( i ) h d

Human (testing) methodsHuman (testing) methodsInspectionW lk h h

http://find/


35/108



Inspection




session.

Leading the session


Author of the product (analyst, designer, programmer)Secretary



H (t ti ) th d

Human (testing) methodsHuman (testing) methodsInspectionW lkth h

http://find/


36/108



Inspection




session.

Leading the session


Author of the product (analyst, designer, programmer)SecretaryReader




Human (testing) methodsHuman (testing) methodsInspectionWalkthroughs

http://find/


37/108



Inspection




session.

Leading the session



Checklists





http://find/


38/108



Inspection




session.

Leading the session



Checklists

Time - 90-120 minutes





http://find/


39/108



Inspection activities

Planning





http://find/


40/108




Planningthe moderator selects the team members;





http://find/


41/108

( g)Questions

gPair-ProgrammingDesk check


Planningthe moderator selects the team members;distribution of the materials to the members;task assignment





http://find/


42/108

( g)Questions

gPair-ProgrammingDesk check



Presentation/Overview - not compulsory





http://find/


43/108

Questions Pair-ProgrammingDesk check




used to present details to the members of the inspection team.



Human (testing) methodsQ i

Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsP i P i

http://find/


44/108





used to present details to the members of the inspection team.Individual preparation



Human (testing) methodsQ ti

Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsP i P i

http://find/


45/108





used to present details to the members of the inspection team.Individual preparationreading and understanding the received documentation;




Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsPair Programming

http://find/


46/108




Presentation/Overview - not compulsoryused to present details to the members of the inspection team.

Individual preparationreading and understanding the received documentation;

Inspection meeting




Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsPair Programming

http://find/


47/108






Inspection meetingcritical observations of each individual inspectors - discussed;




Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsPair-Programming

http://find/


48/108






Inspection meetingcritical observations of each individual inspectors - discussed;conclusions of the inspection - documented;





http://find/


49/108

Questions Pair ProgrammingDesk check






Rework





http://find/


50/108

Qu g gDesk check






Reworkthe author makes the required changes and correct the errors;





http://find/


51/108

g gDesk check







Reinspection/Follow-up





http://find/


52/108

Desk check







Reinspection/Follow-upto verify if the modification did eliminate the errors;





http://find/


53/108

Desk check







Reinspection/Follow-upto verify if the modification did eliminate the errors;may be only between the author and the moderator;




Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsPair-ProgrammingD k h k

http://find/


54/108

Desk check

Inspection checklists

Inspection scope - to find errors




Human (testing) methodsHuman (testing) methodsInspectionWalkthroughsPair-ProgrammingD k h k

http://find/


55/108

Desk check


Inspection scope - to find errorsDepending on the analyzed document - special kind of errors





http://find/


56/108

Desk check


Inspection scope - to find errorsDepending on the analyzed document - special kind of errorsSpecification Document





http://find/


57/108

Desk check



Does the specification conforms to the user’s needs?





http://find/


58/108

Desk check




Are there ambiguities in the specification?





http://find/


59/108

Desk check




Are there ambiguities in the specification?Do the input/output date are clearly stated? What aboutinput/output conditions?





http://find/


60/108



Does the specification conforms to the user’s needs?Are there ambiguities in the specification?Do the input/output date are clearly stated? What aboutinput/output conditions?Are there requirements that are not present in thespecification?





http://find/


61/108



Does the specification conforms to the user’s needs?Are there ambiguities in the specification?Do the input/output date are clearly stated? What aboutinput/output conditions?Are there requirements that are not present in thespecification?Are there performance conditions? What precise computation

conditions?





http://find/


62/108




conditions?Analysis Document





http://find/


63/108





Does the design conforms to the specification?






http://find/


64/108





Does the design conforms to the specification?Are all the functionalities from the specification specified?








65/108





Does the design conforms to the specification?Are all the functionalities from the specification specified?Is there an analysis documentation about the made decisions?








66/108


Code








67/108


Code

Does the code conforms to the design?






http://find/


68/108


Code

Does the code conforms to the design?Are all the methods are called?








69/108


Code

Does the code conforms to the design?Are all the methods are called?Are all the variables initialized?






http://find/


70/108


Code

Does the code conforms to the design?Are all the methods are called?Are all the variables initialized?Problems with: infinite cycles, out of bound indexes, improperallocation of memory.






http://find/


71/108


Code


Test Document






http://find/


72/108


Code


Test Document

The test cases are well documented?






I h kl



73/108


Code


Test Document

The test cases are well documented?The test cases are well chosen?






I i h kli



74/108


Code


Test Document

The test cases are well documented?The test cases are well chosen?Are the test data sufficient to coverage criterion?






I i h kli

http://find/


75/108


Code


Test Document

The test cases are well documented?The test cases are well chosen?Are the test data sufficient to coverage criterion?For the integration testing, the order of integration is clear?






I ti h kli t

http://find/


76/108


Code


Test Document

The test cases are well documented?The test cases are well chosen?Are the test data sufficient to coverage criterion?For the integration testing, the order of integration is clear?At regression testing is the testing continued?






I ti d t



77/108

Inspection advantages

Early error discovery;







http://find/


78/108



Reduce product development time and cost;


OutlineSoftware quality assessmentHuman (testing) methods

Questions




http://find/


79/108




Group method;



Questions




http://find/


80/108




Group method;

Mean to education;



Questions






81/108




Group method;

Mean to education;

The source of error is known;



Questions




http://find/


82/108




Group method;

Mean to education;

The source of error is known;

Eliminates the debugging stress if few day remains until

product release.



Questions



Walkthroughs



83/108

Walkthroughs

Walkthroughs - process of trying to find defects indevelopment documents during various phases of the softwaredevelopment process.

Lect dr Andreea Vescan Software Systems Verification and Validation


Questions



Walkthroughs



84/108

Walkthroughs

Walkthroughs - process of trying to find defects indevelopment documents during various phases of the softwaredevelopment process.Similar to Inspection



Questions



Walkthroughs



85/108

Walkthroughs

Walkthroughs - process of trying to find defects indevelopment documents during various phases of the softwaredevelopment process.Similar to InspectionTeam members ([3-5] members



Questions



Walkthroughs

http://find/


86/108

Walkthroughs


Moderator



Questions



Walkthroughs



87/108

Walkthroughs


ModeratorSecretary



Questions



Walkthroughs

http://find/


88/108

g


ModeratorSecretaryTester



Questions



Walkthroughs

http://find/


89/108

g





Questions



Walkthroughs

http://find/


90/108

g



Procedures are slightly differentPlanning

Meeting - the participants “play computer”.ReworkFollow-up



Questions



Walkthroughs

http://find/


91/108

g





Different error-detection technique



Questions



Walkthroughs

http://find/


92/108





Different error-detection techniqueTime - 90-120 minutes



Questions



Pair-Programming

http://find/


93/108

Variation of program inspection.



Questions



Pair-Programming

http://find/


94/108


Merges coding and inspection activities.





Pair-Programming

http://find/


95/108


Merges coding and inspection activities.The inspection activities





Pair-Programming

http://find/


96/108



are not driven by checklists;





Pair-Programming

http://find/


97/108



are not driven by checklists;are based on shared programming practice and style;

L t d A d V s S ft S st s V ifi ti d V lid ti




Pair-Programming



98/108




L t d A d V S ft S t V ifi ti d V lid ti




Pair-Programming



99/108




Programmers frequently alternate roles;





Pair-Programming

http://find/


100/108




Programmers frequently alternate roles;Is carried out in normal work days, without excessive overtimeand without severe schedule pressure.





Pair-Programming



101/108




Programmers frequently alternate roles;Is carried out in normal work days, without excessive overtimeand without severe schedule pressure.

No mediator, so responsibility for open and nondefensivediscussion of decisions/alternatives falls to the programmers.





Desk check



102/108

The first line of defense against defects.





Desk check



103/108


You can speed up formal inspections by taking care of simpledefects in desk checks first.





Desk check

http://find/


104/108



BUT ...


Outline


Questions

Human (testing) methodsHuman (testing) methods


Desk check

http://find/


105/108



BUT ...

The most basic and least productive type of human-basedtesting.


Outline


Questions

Human (testing) methodsHuman (testing) methods


Desk check

http://find/


106/108



BUT ...

The most basic and least productive type of human-basedtesting.

Many errors are missed in desk checking because it is humannature to overlook your own mistakes;


Outline


Questions

Questions

Questions

http://find/


107/108

Thank You For Your Attention!


Outline


Questions

Questions

Questions

http://find/


108/108

Thank You For Your Attention!

http://find/

Documents

SSVV 01b Inspection