Rapid Application Development and DSDM

1
Rapid Application Development and DSDM

• Slides prepared by S. McRobb and amended by J.
  Ivins

2
Overview of lecture

• This lecture looks at an industry-standard design
  methodology that continues the idea of
  participative design PD
• The model incorporates a lifecycle model and
  explicitly requires user involvement in the
  design process

3
Background

• The emergence of RAD
• DSDM - a methodology for RAD
• The 9 Principles of DSDM
• Critical success factors for RAD projects
• The RAD team
• NB DSDM and RAD were developed by software
  engineers

4
Drivers for RAD

• Sources of frustration / uncertainty
• Business drivers
• Development process drivers
• Examples
• frustration with project timescales
• problems with early requirements freeze
• continued poor involvement of users
• some elements of RAD were already familiar

5
Early days of RAD

• Reactive
• Fragmented
• Often little more than hype
• ( e.g. RAD use of VB )

6
Dynamic System Development Method

• Standard RAD method since 1994
• A Software Engineering Standard
• Industry-wide Consortium
• IBM, ICL, LBMS, British Airways, etc.
• Mission-
• develop and continuously evolve a public domain
  method for RAD

7
DSDM

• 5 phases-
• feasibility
• business study
• functional prototype iteration
• design prototype iteration
• implementation
• Can we apply to multimedia production?

8
Project Selection

• appropriate projects have
• functionality that is visible at the user
  interface
• all users clearly identified
• limited computational complexity
• if large, can be easily split into increments
• a real time-constraint
• flexibility in the detailed requirements

9
DSDM
Lifecycle Model
Feasibility
Business Study
Implementation
Functional model
Design build
10
DSDM9 Principles

• active user involvement
• teams are empowered
• focus on frequent product delivery
• fitness for business purpose is the essential
  criterion for acceptance
• iterative and incremental development

11
9 Principles (continued)

• reversibility of all changes
• requirements baselined at high level
• testing integrated throughout life-cycle
• collaborative and cooperative approach

12
Timeboxing

• absolute time frame for a deliverable
• product-, not activity-based
• unfinished products delivered on time
• deliver what you can.
• . but it must work !

13
Timebox

• Can apply to
• whole project
• single component
• Focuses minds on delivery
• Improves manager client confidence
• Built into project schedule very early
• Not ( usually ) changed later

14
Timebox planning advice

• Short (2-6 weeks optimal)
• Prioritise requirements with MoSCoW
• Must have...
• Should have...
• Could have...
• Want to have...

15
Timebox planning advice
checkpoints
C
O
O
C
C
O
refine
consolidate
investigate
set objectives
16
Managers role

• A logistical challenge
• Move from analysis to design very fast
• No room for slip
• All facilities must be ready when needed
• Little emphasis on change control
• scope controlled by plan and timeboxes
• within timebox, changes are made as easy as
  practicable

17
Managers role

• Project contingency
• .is accounted for by MoSCoW rating of
  requirements
• .thus no contingency in schedule
• Gantt chart usually of little use
• You know where the project has got to by a given
  date!

18
Project communications

• Typically short sharp daily meetings (30 mins
  maximum)
• Focus on delivery, not activity
• BA rule no-one is allowed to say
• I am doing
• Can only say
• I have done...

19
Organisation impact

• Significant culture change
• Education training required
• Project Board workload may increase immensely
• Approval stage much more important, as less
  opportunity to halt / refocus later

20
Summary

• This topic has covered
• Reasons for RAD
• Principles of DSDM
• Implications of DSDM for
• manager
• teams
• organisation

21
Summary continued

• Three models have been introduced in Lecture 3 a
  b.
• Each has strengths and weaknesses
• DSDM appears to offer a model and a methodology
  which can be applied to multimedia design.

22
Reading

• Stapleton J (1997), DSDM, Addison-Wesley (covers
  V2).
• DSDM Consortium at-
• www.dsdm.org