L-1

NEED FOR S/W ENGG

ISSUES IN DESIGN OF LARGE SOFTWARE
S/W LIFE CYCLE MODELS

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

• S/W: in day today life.
• Many application areas: Engg.,architecture,e-commerce,science
• s/w becomes large,complex.(GUI,client-server architecture)
• Run on 2 or more processors, under different OS and on geographical distributed machines.

• Urgent need to adopt s/w engg. concepts,strategies,practices to improve s/w development

process in order to deliver good quality, maintainable s/w in time and within budget.
• S/W process: way in which we produce s/w, differs from organization to organization.
• s/w Process + Management

s/w Engineering:
• Means producing good quality, efficient, maintainable s/w on time, within budget.
• Main focus: On both quality of product and on process used to develop product.
• It’s study of and practice of engineering to build, design, develop and
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
maintain software
• Program vs S/W:
• s/w project development is different from program development that we
develop as a student. We develop it, use it and throw it.
• But now want project-long term + testing + maintenance.

Program Documentation

Operating Procedures

• Documentation: diff. types of manuals(DFD,Flowcharts,ER diagrams etc).

• Operating Procedures: instructions to set up and use software system and
failures(Beginner guide,Installation guide etc).
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
S/w Curve:

Failure Intensity->

-> Time
S/w becomes reliable overtime instead of wearing out.
It may become obsolete, retired due to environmental changes,new requirements,new expectations.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 Need for s/w Engineering
• Basic objective of software engineering is to develop methods and procedures for software development
that can scale up for large systems and that can be used consistently to produce high-quality software at low cost and with a
small cycle of time.

• An Engg. approach consist of developing s/w which meets specifications efficiently, cost effectively and at the same time they
ensure quality.

• Engg. Approach means: well defined, managed, repeatable and predictable approach.
• Large projects will benefit by systematic approach.

• If we are successful once, then same approach, same methodology can also be used in another approach, in another project
and we will have very high probability of success.
• Quality, control,standards,planning and management are important in engineering approach.

Engineering approach: other discipline example:

• Building bridges, dams, power plants, aircraft, missile(involve diff. types of people).
• Big engineering projects: successfully completed.
• we need to learn from them and apply the same for developing s/w.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Need for s/w Engineering continued….
Engineering Approach
1. Attempt to estimate cost/effort/time
2. Plan and schedule work.
3. Involve users in defining requirements.
4. Identify stages in development
5. Define clear milestones so that progress can be measured.
6. Schedule reviews both for control and quality
7. Define deliverables(SRS,SDD).
8. Plan extensive testing.
9. Well-documented maintainable software.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 Issues in design of large s/w:

• Effort intensive( manpower, hardware, software, and the other support resources).

• High cost involved
• Long development time required
• Understanding large and complex system requirements is difficult
The word ‘large’ represents 2 aspects:
(i) Large constraints in terms of security, etc. due to a large number of users.
(ii) Large number of functions to be implemented.

• Customers/Stakeholders are not clear about their needs(happen when they have a very basic idea about their needs)
• Changing needs of users
• Conflicting requirements
• Changing technologies
• High risk of failure and user acceptance
• Performance
• Maintainability: maintaining agility and efficiency
• s/w projects may not always been successful.
• Lack of Resources
• Feasibility
• Undefined system boundaries
There might be no defined set of implementation requirements. Customer may go on to include several unrelated and unnecessary functions besides the
important ones, resulting in an extremely large implementation cost which may exceed the decided budget.
• Partitioning the system suitably to reduce complexity
The projects can sometimes be broken down into small modules or functionalities which are then handled by separate teams. Often, more complex and large
projects require more partitioning. It needs to be ensuredEr.that theRajput,
Shweta partitions are non-overlapping
Programmer, and independent of each other.
UIIT, HPU, Shimla
 Reasons for s/w failures:
1. Schedule –slippage means s/w is not ready in time.
2. Cost over-runs(too costly to complete project, it may be abandoned).
3. Doesn’t solve user’s problem(s/w development completed but user does not find it useful. It’s failure).
4. Poor quality of s/w.
5. Poor Maintainability.

Reasons for such Failures: adhoc, non-engineering approach to s/w development

1. No planning or development work(no milestone defined).
2. Poor understanding of user requirements
3. No control or review(control s/w development systematically and review progress systematically).
4. Technical incompetence of developer.
5. Poor understanding of cost and effort by both developer and user.
6. Outdated technology
7. Security attacks: virus#malware
8. Poor efficiency,performance
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
S/W Life Cycle Models
• Abstraction that represents s/w life cycle.
1. Build and fix
2. Waterfall
3. Prototyping
4. Iterative enhancement model
5. Spiral model
6. RAD(Rapid Application development)

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 Build
1. Build and fix code

fix

• Developer simply build product that’s reworked as many times as necessary to satisfy client.
• Adhoc approach: not well defined.
• It’s 2 phase model:
Phase 1: is to write code.
Phase 2: is to fix it(fixing means error correction or addition of further functionality).
• Suitable for small programming exercises 100 or 200 LOC but not for large s/w.
• Code soon becomes unfixable and unenhanceable.
• Development cost: high as compared to cost of properly specified and carefully designed
product.
• Maintenance can be difficult without specification or design documents.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
2. Waterfall Model
5 phases: top-down approach
1. Requirements analysis and specification
2. Design
3. Implementation
4. unit testing, Integration and system testing
5. Operation and maintenance

Phases always occur in this order and don’t overlap.

Developer must complete each phase before next phase begins.
1st step takes i/p from it’s previous step and gives its output to next step.
• Called so as its diagrammatic representation resembles a cascade of waterfalls.
• Sequential model

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Requirements
analysis and
specification

Design

Implementation
(Coding)
unit testing,
Integration and
system testing

Operation and
maintenance
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
 Phase 1: Requirements analysis and specification

Aim:
1. to understand exact requirements of customer and to document them properly.
2. To develop SRS(software requirement specification) document, User involvement required
to design efficient end product.
3. SRS document: act as contract b/w developer and customer.
written in natural language, contains description of “What system will do without describing
how it will be done?”
4. Techniques: Brainstorming, questionnaire, GD, Survey, Interview etc.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 Phase 2: Design
• Activity following requirements specification and done before programming.
• SRS Document produced in previous phase contains exact requirements of customer.
• Here aim: is to transform SRS into a structure that’s suitable for implementation in some programming language.

• Software design is a process to transform user requirements into some suitable form, which helps the programmer in software coding and
implementation.

• software design may be as simple as a flow chart or text describing a planned sequence of events(pseudocode/algorithm).
• uses design methods like UML(Unified Modeling Language), ER and DFD.
• output of software design process is design documentation, pseudo codes, detailed logic diagrams, process diagrams, and detailed description of all
functional or non-functional requirements.

• Software design : process of defining software methods, functions, objects, and the overall structure and interaction of your code so that the resulting
functionality will satisfy your users requirements. 
• sequence of steps that enables the designer to describe all aspects of the software for building.
• Here, overall s/w architecture is defined and high level and detailed design work is performed and this work is documented in form of SDD(s/w design
document).
• Software design is a creative activity in which you identify software components and their relationships, based on a customer’s requirements.
• if errors get unnoticed till later phases, it becomes more difficult to correct them.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Phase 3: Implementation
• Here design is implemented.
• If SDD is complete, coding phase proceeds smoothly (as all info.
needed by s/w developer is contained in SDD).

• – Software design is a creative activity in which you identify software

components and their relationships, based on a customer’s
requirements.
• – Implementation is the process of realizing the design as a program.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Phase 4: unit testing, Integration and system testing
• During testing, major activities are centred around examination and modification of code.
• Aim of unit testing: is to determine that each independent module is correctly implemented.
• Initially, small modules are tested in isolation from rest of s/w product.
• Effective testing will contribute to delivery of higher quality s/w products, more satisfied
users, lower maintenance cost and more accurate and reliable results.
• Expensive activity( AFTER CODING) and consumes 1/3 to ½ of cost of development project.
• Aim of Integration testing: to check that interface b/w module is also correct.
• System Testing: involve testing of entire system. It is a high level testing always performed
after integration testing.
• conducted on a complete, integrated system to evaluate the system's compliance with its
specified requirements. 

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Phase 5:Operation and maintenance
• Release of s/w inaugurates operational and maintenance phase of life cycle.
• Important and challenging task: that includes error correction, enhancement of
capabilities, deletion of obsolete capabilities and optimization.
• Aim: to preserve s/w value over time.
• This phase may span for 5 to 50 years whereas development may be 1 to 3 years.
4 types of maintenance:
1. Corrective(fixing errors found in day-today system functions)
2. Adaptive (To make s/w adaptable to new environment such as to run the s/w on a new operating system/Hardware/rules)
3. Perfective (adding new functionalities as per the users' changing needs, enhancing both function and efficiency of code)
4. Preventive(assures optimal working conditions and conserves the life span of s/w)
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Corrective maintenance:
• deals with repair of faults or defects or errors(fixing errors) found in day-today system functions, when s/w is in use.
Adaptive maintenance :
• change in software that takes place to make s/w adaptable to new environment such as to run the s/w on a new  operating system.
• changes in the environment such as the hardware or the operating system.
• Term environment in this context refers to conditions and influences which act (from outside) on system. For example, business rules, work
patterns, and government policies have a significant impact on the software system
Perfective maintenance:
• change in the software that occurs while adding new functionalities in the software.
• making functional enhancements to system: To increase the system's performance
• deals with implementing new or changed user requirements.
• This includes enhancing both function and efficiency of code and changing the functionalities of the system as per the users' changing needs.
• Example: modifying payroll program to incorporate a new union settlement and adding a new report in the sales analysis system.
Preventive maintenance
involves implementing changes to prevent the occurrence of errors.
• It tends to reduce the software complexity thereby improving program understandability and increasing software maintainability.
• It comprises documentation updating, code optimization, and code restructuring.
• Documentation updating involves modifying the documents affected by the changes in order to correspond to the present state of the system.
• Code optimization involves modifying the programs for faster execution or efficient use of storage space.
• Code restructuring involves transforming the program structure
Er. Shweta forProgrammer,
Rajput, reducing UIIT,
the HPU,
complexity
Shimla in source code and making it easier to understand.
Waterfall model :
easy to understand based on “define before design ” and “design before code”.

Merits:
1. Useful when requirements are well understood.
2. Same type project: If an organization has experience in developing accounting system then building new
accounting system based on existing design could be easily managed with waterfall model.

Limitations:
3. It expects complete and accurate requirements early in the process which is unrealistic. It assume all
requirements can be clearly defined during analysis phase.
4. Linear model, real projects are rarely sequential.
5. Assume requirements are stable, don’t change with time.
6. Documentation-heavy as too much documentation is being produced as part of project.
7. Working s/w is not available until late in process, so delaying discovery of serious errors.
8. Not covering risk-assessment.
9. Not suitable for accommodating any change(phase sequence can’t be altered).

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 3. Prototyping model
• Disadvantage of waterfall model:
Working s/w is not available until late in process, so delaying discovery of serious errors.
• Here aim is to 1st develop working prototype of s/w instead of developing actual s/w.
• Working prototype is developed as per current available requirements.
• Prototype: has limited functional capabilities, low reliability and untested performance.
• Developers then use this prototype to remove uncertainties in requirements , to refine requirements and
prepare clear, final specification document.
• Since working prototype has been reviewed and evaluated by customer so resulting SRS will be correct.
• Prototype may be usable program but code for prototype is thrown away(or sometimes reworked) as it’s
not suitable as final s/w product.
• Experience gained from developing prototype helps in developing actual system.
• Development of prototype might involve extra cost but overall cost might turn out to be lower than that of
an equivalent system developed using any other model like waterfall.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Aim: to determine customer’s real needs.
Prototype should be developed as early as possible to speed up s/w development process.
• After finalization of SRS document, prototype is discarded and actual system is then
developed using waterfall approach.
• Produces maintainable and good quality s/w
Disadv:
Requires extensive participation and involvement of customer which is not always possible.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 Requirements

Quick Design
(Prototype)

Implement Refinement of requirements as

per suggestions.

Customer Evaluation Not accepted by customer

Accepted by customer

Design

Implementation and
unit testing

Integration and System

Testing

OperationEr.and Maintenence
Shweta Rajput, Programmer, UIIT, HPU, Shimla
4. Iterative Enhancement Model
• Has same phases as waterfall model but with fewer restrictions.
• Phases occur in same order as in waterfall model but these may be conducted in several cycles.
• Useable product is released at end of each cycle, with each release providing additional
functionality.
• During 1st requirement analysis phase, customer and developer prepare SRS document, prioritize
requirements.
• Developer implement specified requirements in 1 or more cycles of design, implementation and
test based on defined priorities.
• Model does deliver an operational quality product at each release but 1 that satisfies only a subset
of customer’s requirements.
• Complete product is divided into releases and developer delivers product release by release.
• 1st release may be available within few weeks or months and customer is able to do some useful
work after 1st release.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

• Here, we define initial scope and try to build product which will meet initial scope that
was identified and then release it quickly to get customer feedback.
• Here we are developing a product whose functionality may not be clearly known or
there may be too many features that we may want to provide. But we would rather
release product quickly, capture the market and get customer feedback. So this is
situation ideal for iterative development.
• here development phase takes place in multiple iterations and many versions of s/w
may be released.
• Latest version represents additional features which have been added from previous
version based on requirements as perceived or based on customer feedback.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 Implementation Integration Operation
Requirements Design And and system (Install)
Unit testing testing Release-1

Implementation
Integration and Operation
Design And
system testing (Install)
Unit testing
Release-2

Implementation Integration and Operation

Design And system testing (Install)
Unit testing
Release-3

Iterative Enhancement Model

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 5. Spiral Model
• Traditional models don’t deal with uncertainty/RISK which is inherent to s/w projects.
• Important s/w projects failed: As project risks were neglected and nobody was prepared when
something happened.
• In this model, project risk factor is incorporated in s/w life cycle.
• It’s a combination of iterative enhancement model and waterfall model.
• It allows incremental releases of product or incremental refinement through each iteration around
spiral.
• Radial dimensions of model represents cumulative costs(refers to the accumulated cost).
• Each path around spiral is indicative of increased costs.
• Angular dimension represents progress made in completing each cycle.
• Each loop of spiral from x-axis clockwise through 360 degree represents 1 phase.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
In spiral model, we go through development in cycle and
each cycle is divided into 4 quadrants(4 sections of major activities).

In 1st quadrant, starting point, we define objectives, alternatives, what are constraints?
Then we evaluate those alternatives and clearly identify the risks.
Based on these risks in 3rd quadrant of cycle, work on implementation, build development and
then in 4th quadrant, we will plan subsequent cycles after customer feedback.
1. Planning: determination of objectives, alternatives and constraints.
2. Risk Analysis: Analyse alternatives and attempts to identify and resolve risks involved.
Risk Analysis includes identifying, estimating, and monitoring technical feasibility and management
risks, such as schedule slippage and cost overrun. 
3. Development: product development and testing product.
4. Assessment: Customer evaluation.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

• During 1st phase,
planning is performed, risks are analysed, prototype are built and customer evaluate prototype.

• During 2nd phase, more refined prototype is built, requirements are documented and validated after refinements and
customer are involved in accessing new prototype.

• By the time 3rd phase begins, risks are known.

• Here focus on identification of problems and classify them into different levels of risks.
• Aim: to eliminate high risk problems before they threaten s/w operation or cost.

• Each phase is completed with review by people concerned with project.

• This review consist of review of all products developed up to that point and include plans for next cycles.
• These plans may include partition of product in smaller portion for development that are implemented by individual groups or
persons.
• If plan for development fails, then spiral is terminated. Else, it terminates with initiation of new or modified s/w.
• So, development is going on in cycles.
• AS we go through cycles, different risk issues are addressed and more refinements are done to our strategies, to the
functionalities.
• Spiral model basically identify risks at every step and it identifies ways of addressing those risks either by following prototyping
methods or carrying out benchmarking.
• If performance are well understood, then risk may be just development risk(go for coding, risk free). Then we follow waterfall
model.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
 Typical uses of Spiral model:
1. budget constraint and risk evaluation is important.
2. For medium to high-risk projects.
3. Good for Long-term project commitment because of potential changes to priorities as
the requirements change with time.
4. Customer is not sure of their requirements which is usually the case.
5. Requirements are complex and need evaluation to get clarity.
6. New product line which should be released in phases to get enough customer
feedback.
7. Significant changes are expected in the product during the development cycle.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

ADV:
1. it’s risk driven. Risk analysis and validation steps eliminate errors in early phases of development.
2. It incorporated s/w quality objectives into s/w development.
3.  It allows for elements of the product to be added in when they become available or known. This
assures that there is no conflict with previous requirements and design.
4. It is consistent with approaches that have multiple software builds and releases and allows for
making an orderly transition to a maintenance activity.
5. Another positive aspect is that the spiral model forces early user involvement in the system
development effort.

Disadvantages of Spiral model:

• Difficult to follow strategy for small projects.
• Not much useful for low risk projects.
• Need more experience resources as process is bit complex.
• Large documentation.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
 6. RAD(Rapid Application Development)
• By IBM
• Here user involvement is essential from requirement phase to delivery of product.
• Continuous user participation: in preparation of SRS and SDD documents.
• Firstly rapid prototype is designed and given to user for evaluation.
• User feedback is obtained and prototype is refined.
• Process continues till requirements are finalized.
• SRS and SDD documents are prepared with association of users.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

 With active participation of users

Requirements Construction
User description Cut-Over
Planning

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

1. Requirements Planning
Here requirements are captured using any group elicitation technique. User involvement is necessary
for understanding the project.

2. User description:
Joint team of developers and users are constituted to prepare, understand and review requirements.
Team may use automated tools to capture information from other users.

3. Construction:
It combines detailed design, coding and testing phase of waterfall model. Here we release product to
customer. It may use code generators, screen generators and other types of productivity tools.

4. Cut-Over:
It incorporates acceptance testing by users, installation of system and user training.
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Advantages:
1. User involvement may increase acceptability of product.
2. Development time of product may be reduced due to use of
powerful development tools. It may use CASE tools and framework
to increase productivity.
3. Quick initial views about product are possible due to delivery of
rapid prototype.
4. Highly specialised and skilled developers are expected.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

• Selection of suitable model is based on foll. characteristics:

1. Requirements
2. Development team
3. Users
4. Project type and associated risks.

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Requirements

1. If requirement are easily defined and understandable?

Waterfall, RAD

2. Change requirements
Prototype, Spiral

3. Requirements indicate that system is complex.

Prototype, Spiral, iterative enhancement model

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Development team

4.If development team has less experience on similar project.

Prototype, Spiral

5. If development team has less domain knowledge(new technology)

spiral, iterative enhancement

6. If development team has less experience on tools to be used.

Spiral
Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Users
7. User involvement in all phases.
Prototype, RAD ,spiral

8. Limited user participation.

Waterfall, iterative enhancement model

9. User have no previous experience of participation in similar project.

Prototype, spiral, iterative enhancement model

10. Users are expert of problem domain.

RAD, Prototype, iterative enhancement model

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla

Project type and associated risks.
11. Project is enhancement of existing system
RAD, iterative enhancement model

12. Funding is stable for project

waterfall, prototype, RAD

13. High reliability requirements

spiral, iterative enhancement model

14. Tight project schedule:

spiral, iterative enhancement model, RAD, Prototype

15. Use of reusable components.

Prototype,spiral,RAD

16. Resources(time,money,people etc) less?

Er. Shweta Rajput, Programmer, UIIT, HPU, Shimla
Prototype, spiral