Rugby - A Process Model for Continuous Software Engineering

[email protected] - www.skrusche.de - @skrusche
Dissertation | Stephan Krusche | March 11, 2016
Rugby - A Process Model for
Continuous Software Engineering

View Slide

Dissertation | Stephan Krusche | Rugby - A Process Model for Continuous Software Engineering
“In reality, software development is a complex, continuous, iterative, and repetitive
process” — Carolyn Wong, 1984
2
Deﬁned Process  
Control Model
Empirical Process  
Control Model
“When the process is too complicated for the deﬁned approach, the empirical
approach is the appropriate choice” — Babatunde Ogunnaike, 1994
• Entirely planned
• Follows strict rules
• Treats deviations as 
problems that need  
to be resolved
• Not entirely planned
• Inspect and adapt
• Treats deviations as  
opportunities that  
lead to improvements

View Slide

Deﬁned Process Control
3
Developer
Requirements
Process
System
Allocation
Process
Concept
Exploration
Process
Design
Process
Implementation
Process
Installation
Process
Operation &
Support Process
Verification
& Validation
Process
[Royce 1970]
Traditional Software Engineering Software Use
User
Educator
Traditional Lecture Exercise
Student
Delay (months / years)
Delay (weeks)

View Slide

Empirical Process Control
4
Delay (weeks)
Delay (days)
Software Usage
User
Blended Learning
Experiential Learning [Kolb 1984]
[Garrison and Kanuka 2004]
Educator
Exercise
Student
Agile Software Engineering
Developer
[Schwaber 1995]

View Slide

• Process Model for Continuous Software Engineering that addresses the delay in
software projects
• Extensible to address the delay in education —> Continuous Education
• Frequent event based interactions can reduce difﬁculty and complexity
• Hypotheses:
H1) Rugby allows to reduce the delay between development and usage in software projects
H2) Rugby allows to reduce the delay between lectures and exercises in education
Rugby: a new Process Model
5

View Slide

• Develop and release small changes to software in fast, parallel cycles  
in order to obtain user feedback and incrementally improve software [Bosch 2014]
• Event based releases allow to reduce the delay between 
development and usage to minutes / hours
• Multiple, parallel, small iterations in one Sprint
• Each release leads to a potential software increment
6
User
Usage
Release
Feedback
Development
Adaption

View Slide

Continuous Education
• Teach and exercise small content in fast cycles so that students reﬂect about it
immediately and increase their knowledge incrementally
• Combine lectures and exercises into an interactive session 
to increase the knowledge retention
• Multiple, small iterations in one session
• Each exercise leads to a potential knowledge increment
7
Exercise
Example
Solution
Student
Reﬂection
Theory

View Slide

Rugby’s Process Meta Model
• Rugby is based on a process meta model that treats changes as events in parallel
running workﬂows
• Changes can occur at any time in the development process
8

View Slide

Rugby’s Process Meta Model Hierarchy (Software Engineering)
9
M3 - Meta Meta Model
Meta Object Facility  
[OMG 2005]
M2 - Meta Model
Generic concepts
M1 - Model
Development methodology
Process Model
M0 - Reality
What actually happens
Concrete Development
Process
Class
«stereotype»
Work Queue
«instantiate»
«Work Queue»
Sprint Backlog
«instantiate»
«instantiate»
«stereotype»
Work Item
«Work Item»
Backlog Item
«instantiate»
«instantiate»
*
*
«instantiate»
Sprint 1 Backlog
ID Name Difﬁculty
1 Search available Pedelecs Medium
2 Check working radius Large
«instantiate»

View Slide

Rugby’s Process Meta Model Hierarchy (Education)
10
M3 - Meta Meta Model
Meta Object Facility  
[OMG 2005]
M2 - Meta Model
Generic concepts
M1 - Model
Teaching methodology
Process Model
M0 - Reality
What actually happens
Concrete Teaching  
Process
Class
«stereotype»
Work Queue
«instantiate»
«Work Queue»
Exercise Backlog
«instantiate»
«instantiate»
«stereotype»
Work Item
«Work Item»
Exercise
«instantiate»
«instantiate»
*
*
«instantiate»
Exercise 1 Backlog
ID Name Difﬁculty
1 Bubble Sort Easy
2 Quick Sort Hard
«instantiate»

View Slide

Workflow Meta Model
Change Meta Model
-load
-state : State
+customize()
+activate()
+sleep()
Workflow
+modifyItem()
+proritizeItem()
Work Queue
Configuration Management Meta Model
-state : State
-priority : Int
-assignee : Team Member
+interrupt()
+resume()
Work Item
Artifact
(Resource)
Role Meta Model
Sleep
Ready
Running
Blocked
Finished
< < e n u m > >
State
Team
Member
Organization
+scheduleEvent()
+reactToUnscheduledEvent()
+giveFeedback()
Role
Activity
+analyze()
+design()
+implement()
+test()
Developer
+tailorProcess()
+customizeWorkflow()
+scheduleActivity()
+prioritizeActivity()
Project Manager
Project Management Meta Model
+tailor()
Process
Portfolio
Project
Outcome
Controlled
Item
Educational
Unit
+send()
Event
+addObserver()
+removeObserver()
+notify()
Event Subject
+update()
Event Observer
Service
(Tool)
Feedback
Task
Schedule
Version
Milestone
Phase
Review
Release
Team
+release()
Release Manager
+review()
Reviewer
Organizational
Unit
Iteration
+customizeWorkflow()
Workflow Manager
* *
*
0..*
*
1
*
*
*
1..*
1..*
*
*
*
*
1
*
*
1..*
*
Rugby
Receive Event
Rugby Send
Event
has
works on
depends on
manages
requires
creates
relates to
activates
subscribe
/ create
11

View Slide

Workﬂow Lifecycle
12
Active Finished
Blocked
activate sleep
stop
resume
interrupt
Sleep
End
Start
Rugby
Receive Event
Rugby Send
Event

View Slide

Interaction between two workﬂows
13
Analysis
Design
New
Requirement
Design
Change
New
Requirement
Active Finished
Blocked
activate sleep
stop
resume
interrupt
Sleep
End
Start
Rugby Send
Event
Feature
Change
Architecture
Change
Implementation
Change
Rugby
Receive Event
Design Change

View Slide

Dissertation | Stephan Krusche | Rugby - A Process Model for Continuous Software Engineering 14
Requirements
Elicitation
Analysis
Design
Implementation
Test
Deployment
Time

View Slide

Dynamic View on Rugby’s Workflow Meta Model
15
Analysis
Requirements
Elicitation
Analysis
Design
Implementation
Deployment
Test
Requirements
Elicitation
Design Implementation
Time
Test Deployment
Waterfall Model
Elaboration
Inception
Requirements
Analysis & Design
Implementation
Deployment
Configuration &
Change Management
Test
Business Modeling
Construction
Project Management
Environment
I1 E1 E2 C1 C2 C3 C4 T1 T2
Transition
Time
Unified Process
Sprint 1 … Sprint n
Sprint 2
Time
Analysis
Design
Implementation
Test
Requirements
Elicitation
Project
Management
Scrum
Top Level Design
Problem
Statement
Analysis
Design
Implementation
Project Management
Review Management
Test
Requirements Elicitation
Schedule Kickoff
Release Management
Feedback Management
Product Backlog
Time
Elaboration
Inception
Sprint 0 Sprint 1 Sprint n
…
…
Construction
Event
Document Event based release
Time based release
Key:
Rugby
Analysis
Design
Feedback
…

View Slide

Development
Review
Management
Release
Management
Feedback
Management
Analyze
finished?
Backlog Item ready
for development
Merge Request
Merge Request Review
Quality ok?
yes
yes
Backlog Item finished
Merge
need feedback?
no
yes
no
Improvement Request
Improvement Request
Release Request Release Event based Release
no
Feedback Report
Analyze
Feedback
Is new
requirement?
New Backlog Item
Change Request
yes
no
Change Request
Design
Implement
Test
Rugby integrates Review, Release and Feedback Management Workflows
Release Request
Change Request
Merge Request
Release Request
Improvement Request
Feedback Report

View Slide

Event based Releases in Rugby
17
Sprint End
Master Branch
Feature Branch 1
Development Branch
Sprint Start
Developer User
Unreleasable build
Releasable build
Event based Release
Key:
Realized feature build
Feature Branch 2
Time based Release
[Krusche et. al 2014]

View Slide

User
Usage
Release
Feedback
Development
Adaption
Exercise
Example
Solution
Student
Reﬂection
Theory
3 Case Studies
18
Continuous Software Engineering Continuous Education
Case Study 1:  
62 course  
projects
Case Study 3:  
8 projects  
in industry
Case Study 2:  
5 interactive  
sessions 
in a lecture

View Slide

Case Study 1: Capstone Course Projects
• 62 projects between 2011 and 2015  
with ~ 500 students
• Project duration: 3 months
• H1: Rugby allows to reduce the delay
between development and usage in
software projects
Findings
Metrics per team in 2014:
• 96 code reviews (every 5 hours)
• 64 releases (every 8 hours)
• 135 downloads (every 4 hours)
• 27 feedback reports (every 19 hours)
➡ Reduced delay between development
and usage from 3 months to 8 hours
on average
[Krusche and Alperowitz 2014], [Bruegge, Krusche and Alperowitz 2015]

View Slide

Case Study 2: Lecture
20
Findings
• 57 team projects with ~ 400 students
• 5 interactive sessions (lecture and exercise
workﬂows intertwined) with 26 knowledge
increments
between lectures and exercises in education 
(thereby, the knowledge retention increases)
0%
20%
40%
60%
80%
100%
Improved skills in
interactive sessions
Conﬁdent to apply
the knowledge gained  
in interactive sessions
5%
4% 12%
18%
83%
78%
Agree Neutral Disagree
2
2.5
3
3.5
4
2.4
2.5
2.9
3.6
3.9
GPA (Grade point average, lower is better)
Exercise participation
GPA
20 % 40 % 60 % 80 % 100 %
0 %
➡ Reduced delay between lectures and
exercises to 10-15min on average

View Slide

Case Study 3: Industry
21
Findings
• Reduced time effort for integration and
delivery from several hours to 5 minutes
• Increased frequency of releases
➡ Reduced delay between development
and usage
• 8 projects with 31 professionals tailored
Rugby to their needs
between development and usage in
software projects
[Klepper, Krusche et. al 2015]

View Slide

Contributions
• Tailorable process model for continuous software engineering
• Extensible process meta model that can be applied to software engineering and
education
• 3 large case studies in capstone course, lecture and industry
• Empirical validation of Rugby
Results
➡ Rugby reduces the delay between development and usage in software projects
➡ Rugby reduces the delay between teaching and exercising in education
22

View Slide

[email protected] - www.skrusche.de - @skrusche
Dissertation | Stephan Krusche | March 11, 2016
Rugby - A Process Model for
Thank You!

View Slide

Bibliography
24
[Wong 1984] Carolyn Wong. A successful software development. IEEE Transactions on Software
Engineering, pages 714–727, 1984.
[Ogunnaike and Ray 1994] Babatunde Ayodeji Ogunnaike and Willis Harmon Ray. Process Dynamics, Modeling,
and Control, volume 1. Oxford University Press, 1994.
[Royce 1970] Winston Royce. Managing the development of large software systems. In
Proceedings of IEEE WESCON, 1970.
[Schwaber 1995] Ken Schwaber. Scrum development process. In Proceedings of the OOPSLA
Workshop on Business Object Design and Information, 1995.
[Garrison and Kanuka 2004] Randy Garrison and Heather Kanuka. Blended learning: Uncovering its
transformative potential in higher education. The internet and higher education, 2004.
[Kolb 1984] David Kolb. Experiential learning: Experience as the source of learning and
development, volume 1. Prentice Hall, 1984.

View Slide

Bibliography
25
[Bosch 2014] Jan Bosch. Continuous Software Engineering. Springer, 2014.
[OMG 2005] Object Management Group. Meta Object Facility 1.4.1, International Standard. 2005.
Retrieved January 08, 2016 from http://www.omg.org/spec/MOF
[Krusche et. al 2014] Stephan Krusche, Lukas Alperowitz, Bernd Bruegge, and Martin Wagner. Rugby: An
agile process model based on continuous delivery. In Proceedings of the 1st
International Workshop on Rapid Continuous Software Engineering, pages 42–50.
ACM, 2014.
[Krusche and Alperowitz 2014] Stephan Krusche and Lukas Alperowitz. Introduction of Continuous Delivery in Multi-
Customer Project Courses. In Companion Proceedings of the 36th International
Conference on Software Engineering, pages 335–343. IEEE, 2014.
[Bruegge et. al 2015] Bernd Bruegge, Stephan Krusche, and Lukas Alperowitz. Software engineering
project courses with industrial clients. ACM Transactions on Computing Education,
15(4):17:1–17:31, 2015.
[Klepper et. al 2015] Sebastian Klepper, Stephan Krusche, Sebastian Peters, Bernd Bruegge, and Lukas
Alperowitz. Introducing continuous delivery of mobile apps in a corporate
environment: A case study. In Proceedings of the 2nd International Workshop on
Rapid Continuous Software Engineering, pages 5–11. IEEE/ACM, 2015.

View Slide

My relevant publications
• Bernd Bruegge, Stephan Krusche, and Martin Wagner. Teaching Tornado: from communication models to releases. In Proceedings of the 8th
edition of the Educators’ Symposium, pages 5–12. ACM, 2012.
• Stephan Krusche and Lukas Alperowitz. Introduction of Continuous Delivery in Multi-Customer Project Courses. In Companion Proceedings of the
36th International Conference on Software Engineering, pages 335–343. IEEE, 2014.
• Stephan Krusche, Lukas Alperowitz, Bernd Bruegge, and Martin Wagner. Rugby: An agile process model based on continuous delivery. In
Proceedings of the 1st International Workshop on Rapid Continuous Soft- ware Engineering, pages 42–50. ACM, 2014.
• Stephan Krusche and Bernd Bruegge. User feedback in mobile development. In Proceedings of the 2nd International Workshop on Mobile
Development Lifecycle, pages 25–26. ACM, 2014.
• Bernd Bruegge, Stephan Krusche, and Lukas Alperowitz. Software engineering project courses with industrial clients. ACM Transactions on
Computing Education, 15(4):17:1–17:31, 2015.
• Sebastian Klepper, Stephan Krusche, Sebastian Peters, Bernd Bruegge, and Lukas Alperowitz. Introducing continuous delivery of mobile apps in a
corporate environment: A case study. In Proceedings of the 2nd International Workshop on Rapid Continuous Software Engineering, pages 5–11.
IEEE/ACM, 2015.
• Han Xu, Stephan Krusche, and Bernd Bruegge. Using software theater for the demonstration of innovative ubiquitous applications. In Proceedings
of the 10th Joint Meeting on Foundations of Software Engineering, pages 894–897. ACM, 2015.
• Stephan Krusche, Mjellma Berisha, and Bernd Bruegge. Teaching Code Review Management using Branch Based Workﬂows. In Companion
Proceedings of the 38th International Conference on Software Engineering. IEEE, 2016. Accepted for Publication.
• Dora Dzvonyar, Stephan Krusche, Rana Alkadhi and Bernd Bruegge. Context-Aware User Feedback in Continuous Software Evolution.
International Workshop on Continuous Software Evolution and Delivery, ACM, 2016. Accepted for Publication.
• Sebastian Klepper, Stephan Krusche and Bernd Bruegge. Semi-Automatic Generation of Audience-Speciﬁc Release Notes. International Workshop
on Continuous Software Evolution and Delivery, ACM, 2016. Accepted for Publication.
26

View Slide

Rugby - A Process Model for Continuous Software Engineering

Rugby - A Process Model for Continuous Software Engineering

Stephan Krusche

More Decks by Stephan Krusche

Other Decks in Research

Featured

Transcript