Scribd
Upload
124 views

Software Complexity

Software systems are inherently complex due to factors such as complex problem domains, difficulties managing large development teams, flexibility of software, and challenges characterizing discrete system behavior. Well-designed complex software exhibits attributes like hierarchical structure, separation of concerns, and combination of common patterns. Object-oriented design helps manage complexity by capturing common features and behaviors within classes and illustrating how objects collaborate.

Uploaded by

gevanson967
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
124 views

Software Complexity

Software systems are inherently complex due to factors such as complex problem domains, difficulties managing large development teams, flexibility of software, and challenges characterizing discrete system behavior. Well-designed complex software exhibits attributes like hierarchical structure, separation of concerns, and combination of common patterns. Object-oriented design helps manage complexity by capturing common features and behaviors within classes and illustrating how objects collaborate.

Uploaded by

gevanson967
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 23

SOFTWARE COMPLEXITY

INHERENT COMPLEX NATURE OF SOFTWARE


SYSTEM
COMPLEXITY
Software systems are complex artifacts
 Failure to master this complexity results in projects
that are
 late

 go over budget

 do not meet requirements


COMPLEXITY
The physical world is full of complex systems (both
natural and man-made)

Software’s complexity is however fundamentally


different:

Software is unbound by physical constraints

Industrial(-strength) software exhibits a rich set of


behaviors
INDUSTRIAL SOFTWARE
Process Control (oil, gas, water, ...)

Transportation (air traffic control, ...)

Health Care (patient monitoring, device control, ...)

Finance (automatic trading, bank security, ...)

Defense (intelligence, weapons control, ...)

Manufacturing (precision milling, assembly, ...)


Why Software Is
Inherently Complex
1.Complexity of problem domain
2.Difficulty of managing development
process
3.Flexibility afforded by software
4.Difficulty of characterizing discrete
system behavior
Complexity of
Problem Domain
Many , often contradictory, requirements

 functional(what must be done)

 non-functional(usability, cost, performance,


consumption,...)

Communication gap between customers and developers

Evolving requirements
Difficulty of Managing
Development Process

Fundamental task of software development:


engineer the illusion of simplicity
However, ...
Difficulty of Managing
Development Process
Modern systems are huge (106 LOC, 102 modules)

Development team is necessary

More developers =>

 More complex communication

 More difficult coordination

 harder to maintain design unity/integrity


Flexibility Afforded
by Software
 Software is the ultimate flexible product

 It is technically possible for any developer to create anything with


it

 This is both a blessing and a curse

 Other industries have specialization, codes and quality standards

 Software development remains a mostly artisanal labor-intensive


business
Difficulty of Characterizing Discrete
Systems Behavior

Physical (analog) systems exhibit continuous behavior

Small external perturbations produce small changes in


behavior

Software (digital) systems exhibit discrete behavior

Small changes in input can produce large changes in


output
Difficulty of Characterizing
Discrete Systems Behavior

Discrete systems have a combinatorial state explosion

Describing their behavior precisely and formally is very


challenging in general

Most software professionals are poorly trained for that

Testing for flaws is intrinsically insufficient


Common Features of
Good Complex Systems
1.Nearly decomposable, hierarchic structure
2.Primitive components
3.Separation of concerns
4.Combination of common patterns
5.Stable intermediate forms
Organized complexity
Many hierarchies can be found in a complex system

Most important for us:


object structure ("part of" relation)

class structure ("is a" relation)


We refer to them together as the system's architecture
Canonical Form of
a Complex System
Classes capture common features of a set of objects

Each object is an instance of a class

Objects are composed of and interact with other


objects
Successful Complex
Software Systems
Exhibit the 5 attributes characterizing good complex
systems
Have well designed and built class
and object structures
captures common features and behavior within a
system
illustrates how different objects
collaborate with one another
The Software Development Predicament

The complexity of software systems is ever increasing

The human ability to cope with complexity is


fundamentally limited

Time-honored technique to master complexity:


divide et impera
Decomposition and abstraction are key
Successful Complex
Software Systems
Exhibit the 5 attributes characterizing good complex
systems
Have well designed and built (i) class
and (ii) object structures

captures common features and behavior within a system

illustrates how different objects collaborate with one


another
Alternative
Decomposition Approaches

1. Algorithmic Decomposition: Each component denotes


a major step in the system's overall process
2. Object-Oriented Decomposition: Each component
denotes a major agent in the system's overall process
Main Advantages of
OO Decomposition
It facilitates
 reuse of components and mechanisms

 system evolution over time


 Separation of concerns
The Role of Design in
Software Development
 Construct a system that (Mostow):
 Satisfies a given (perhaps informal) functional specification

 Conforms to limitations of the target medium

 Meets implicit or explicit requirements on performance and resource


usage

 Satisfies implicit or explicit design criteria on the form of the artifact

 Satisfies restrictions on the design process itself, such as its length or


cost, or the tools available
The Importance of
Model Building in Design
Widespread in all engineering disciplines

Appeals to the principles of abstraction decomposition and


hierarchy

Models can be evaluated and modified before the actual


system is built

allow us to focus on important aspects by abstracting away


irrelevant details
Basic Elements of Software
Design Methodologies

Notation: The language for expressing


models
Process: The activities leading to the orderly
construction of a system's model
To o l s: The artifacts that facilitate the creation and
validation of models
Effective OO Design
and Development
 Requires us to master these underlying principles:
 abstraction

 encapsulation

 Modularity

 hierarchy

 typing

 concurrency

 persistence

You might also like