SlideShare a Scribd company logo

PhD thesis defense

Download as KEY, PDF
••
Damien Cassou, profile picture
Damien Cassou

A good software design is crucial for ensuring long-term success. A design framework provides guidance by establishing a design language and paradigm. This framework guides architects and improves collaboration, productivity, and maintenance of the system.

Technology
Related topics:
Software Architecture
1 of 112
Downloaded 19 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
Leveraging Software Design to Guide the Development of Sense/Compute/Control Applications Damien Cassou
Design is Crucial “The most important ingredient in ensuring a soft ware system’s long-term success is its design ” ICSE’11 c.f.p • A good design improves n1155: lf n162 ForLoop n536 start() n1704 n822 • collaboration ArcCond ArcSon ArcCond ArcCond n324 n1303 [name > len] n1304: Block Block error_mark visb: string = unk run_arch() ArcCond ArcCond ArcSon[2] [name <= len] stop_arch() n1702: Operator n1834: Operator • productivity op: string = fct_call ArcSon[1] op: string = fct_call n64: Literal ArcOpd[1] ArcOpd[2] value: string = 0 <<StartState . ActionState >> n7848 success << ActionState >> Suggest Itineraries NameRef n7848 n1966 RefersTo RefersTo NameRef ualFctParam error success n7840 RefersTo n7820: NameRef << EndState >> << EndState >> • maintenance NameRef value: string = len << ViewState >> startOver Display Itineraries RefersTo Try Again n471: Object success Cancel name: string = done n391: Function visb: string = unk name: string = exit static: bool = false << ActionState >> visb: string = pub extern: bool = false Select Itineraries virtual: string = non-virtual const: bool false assign book Instance RefersTo RefersTo n27: BuiltinType << SubFlowState >> finish << ActionState >> name: string = void n7828: NameRef n7846: NameRef Choose Itineraries Book Itineraries visb: string = unk name: string = done name: string = done success Confirmation 2
Design Framework A good design requires a design framework which guides the architect with • a language n536 n1704 n822 start() n1155: lf n162 ForLoop ArcCond ArcSon ArcCond ArcCond n324 • a paradigm n1303 [name > len] n1304: Block Block error_mark visb: string = unk run_arch() ArcCond ArcCond ArcSon[2] [name <= len] stop_arch() n1702: Operator n1834: Operator op: string = fct_call ArcSon[1] op: string = fct_call n64: Literal ArcOpd[1] ArcOpd[2] value: string = 0 <<StartState . ActionState >> n7848 success << ActionState >> Suggest Itineraries NameRef n7848 n1966 RefersTo RefersTo NameRef ualFctParam error success n7840 RefersTo n7820: NameRef << EndState >> << EndState >> NameRef value: string = len << ViewState >> startOver Display Itineraries RefersTo Try Again n471: Object success Cancel name: string = done n391: Function visb: string = unk name: string = exit static: bool = false << ActionState >> visb: string = pub extern: bool = false Select Itineraries virtual: string = non-virtual const: bool false assign book Instance RefersTo RefersTo n27: BuiltinType << SubFlowState >> finish << ActionState >> name: string = void n7828: NameRef n7846: NameRef Choose Itineraries Book Itineraries visb: string = unk name: string = done name: string = done success Confirmation 3
Programming Framework A good implementation requires a programming framework which guides the developer with • abstractions • services 4
Conformance An implementation must conform to its design n536 n1704 n822 start() n1155: lf n162 ForLoop ArcCond ArcSon ArcCond ArcCond n324 n1303 [name > len] n1304: Block Block error_mark visb: string = unk run_arch() ArcCond ArcCond ArcSon[2] [name <= len] stop_arch() n1702: Operator n1834: Operator op: string = fct_call ArcSon[1] op: string = fct_call n64: Literal ArcOpd[1] ArcOpd[2] value: string = 0 <<StartState . ActionState >> n7848 success << ActionState >> Suggest Itineraries NameRef n7848 n1966 RefersTo RefersTo NameRef ualFctParam error success n7840 RefersTo n7820: NameRef << EndState >> << EndState >> NameRef value: string = len << ViewState >> startOver Display Itineraries RefersTo Try Again n471: Object success Cancel name: string = done n391: Function visb: string = unk name: string = exit static: bool = false << ActionState >> visb: string = pub extern: bool = false Select Itineraries virtual: string = non-virtual const: bool false assign book Instance RefersTo RefersTo n27: BuiltinType << SubFlowState >> finish << ActionState >> name: string = void n7828: NameRef n7846: NameRef Choose Itineraries Book Itineraries visb: string = unk name: string = done name: string = done success Confirmation 5
Requirements 1. A design framework to guide the architect  2. A programming framework to guide the developer  3. A guaranteed conformance of the implementation relatively to the design Conformance 6
Related Works Design  Implementation  Conformance Conformance 7
Related Works GPL Design  Implementation +  Conformance Conformance Java 8
Related Works GPL Library Design  Implementation + ++  Conformance Conformance Spring 9
Related Works GPL Library ADL Design ++  Implementation + ++  Conformance Conformance C2 10
Related Works GPL Library ADL ADL++ Design ++ +  Implementation + ++ +  Conformance + Conformance ArchJava 11
Thesis A paradigm-oriented framework for both design and implementation which maintains conformance all along the software life-cycle 12
The Paradigm Sense/Compute/Control (SCC) “applications that interact with an environment” Environment 13
The SCC Paradigm sense compute Environment control 14
The SCC Paradigm sense GPS, flight plan compute compute direction control control aileron, engine 15
The SCC Paradigm motion detection sense compute Environment intrusion? control trigger alarms 16
The SCC Paradigm Covers various domains • pervasive computing • tier-system monitoring Environment • avionics • robotics • ... 17
Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 18
Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 19
Design Language Environment orders 20
Design Language sources Environment • sources sense the environment orders 20
Design Language sources Environment actions • sources sense the environment • actions impact the environment orders 20
Design Language a concept for handling the interaction with the environment sources Environment actions • sources sense the environment • actions impact the environment orders 20
Design Language entity Keypad { source keycode as Integer; a concept for handling action UpdateSt;  } the interaction with the environment action UpdateSt { updateStatus(message as String); } sources Entities Environment actions • sources sense the environment • actions impact the environment orders 20
Design Language entity Keypad { source keycode as Integer; a concept for handling action UpdateSt;  } the interaction with the environment action UpdateSt { updateStatus(message as String); } sources Entities Environment actions 1 entity description for potentially many implementations orders 21
Design Language entity Keypad { source keycode as Integer; a concept for handling action UpdateSt;  } the interaction with the environment action UpdateSt { updateStatus(message as String); } sources Entities Environment actions 1 entity description for potentially many instances orders 22
Design Language entity Keypad { source keycode as Integer; action UpdateSt;  attribute room as Integer; } action UpdateSt { updateStatus(message as String); } sources Entities Environment actions 1 entity description for potentially many instances ➡ attributes to characterize instances orders (color, location, reliability, etc.) 23
Design Language raw data a concept to refine raw data sources Entities Environment actions orders 24
Design Language raw data a concept to refine raw data context operators sources refined information Entities Environment actions context BuildingSecured as Boolean { source keycode from Keypad; } orders  24
Design Language raw data a concept to take decisions based on application-level data context operators sources refined information Entities Environment actions orders 25
Design Language raw data a concept to take decisions based on application-level data context operators sources refined information Entities Environment actions control operators controller BuildingManager { context BuildingSecured; orders orders action UpdateSt on Keypad; }  25
Design Language raw data The language features concepts Compute dedicated to the SCC paradigm context Sense operators sources refined information Entities Environment actions control operators orders orders Control 26
Application logic Environment handling context operators sources refined information Entities Environment actions control operators orders 27
context operators Information creation refined information control operators Information use orders 28
Design Language raw data context operators sources refined information Entities Environment actions control operators orders orders 29
Design Language actions control operators context operators orders sources 30
Design Language actions control operators context operators orders sources 31
Case Study: Anti-Intrusion actions control operators context operators sources 32
Case Study: Anti-Intrusion actions control operators Intrusion context operators sources 32
Case Study: Anti-Intrusion Alarm actions OnOff Intrusion control Manager operators Intrusion context operators sources 32
Case Study: Anti-Intrusion Alarm actions OnOff Intrusion control Manager operators Intrusion context operators Building Presence Secured sources 32
Case Study: Anti-Intrusion Alarm actions OnOff Intrusion control Manager operators Intrusion context operators Building Presence Secured keycode motion Keypad MotionSensor sources 32
Case Study: Anti-Intrusion Keypad Alarm actions UpdateSt OnOff Security Intrusion control Manager Manager operators Intrusion context operators Building Presence Secured keycode motion Keypad MotionSensor sources 32
Case Study: Anti-Intrusion Keypad Alarm Mailer actions UpdateSt OnOff Send Security Intrusion control Manager Manager operators Intrusion context operators Building Scene Presence Secured Image keycode motion image Keypad MotionSensor Camera sources 32
Case Study: Anti-Intrusion Keypad Alarm Mailer actions UpdateSt OnOff Send Security Intrusion control Manager Manager operators Intrusion context operators Building Scene Presence Secured Image keycode motion image Keypad MotionSensor Camera sources 33
Case Study: Anti-Intrusion Intrusion Building Presence Secured 34
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured what the architect has in mind 34
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured what the architect has in mind 34
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured what the architect has in mind 34
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured the architect should express what he has in mind 35
Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured the architect should express what he has in mind  ➡interaction contracts 35
Interaction Contracts context operator 36
Interaction Contracts request 1 Activation condition 1 context operator 36
Interaction Contracts 1 Activation condition context operator event 1 36
Interaction Contracts 1 Activation condition 2 Data requirement context operator event 1 2 request 2 request source context Entity operator 36
Interaction Contracts 1 Activation condition 3 event 2 Data requirement context operator 3 Emission event 1 2 request 2 request source context Entity operator 36
Interaction Contracts 1 Activation condition 2 Data requirement Intrusion 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { when provided Presence get BuildingSecured maybe publish } } 37 
Interaction Contracts 1 Activation condition 2 Data requirement Intrusion 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence get BuildingSecured maybe publish } } 37 
Interaction Contracts 1 Activation condition 2 Data requirement Intrusion 2 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence 2 get BuildingSecured maybe publish } } 37 
Interaction Contracts 1 Activation condition 3 2 Data requirement Intrusion 2 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence 2 get BuildingSecured 3 maybe publish } } 37 
Interaction Contracts 1 Activation condition 3 2 Data requirement Intrusion 2 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence 2 get BuildingSecured 3 maybe publish related to automata } approaches } 37 
Summary  A design framework consisting of a design language, DiaSpec, which guides the architect by offering • concepts dedicated to the SCC paradigm • a separation between environment handling and logic • a separation between information creation and use • a dedicated description of interactions 38
Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 39
A Programming Framework Generated from the Design DiaSpec GPL         40
A Programming Framework Generated from the Design DiaSpec GPL         • separates 2 different roles with 2 different languages • leverages GPL tools, libraries and expertise • ensures conformance automatically 40
A Programming Framework how to make a programming framework conform to a particular design? 41
A Programming Framework The code generator maps • each description to an abstract class • each interaction contract to an abstract method 42
A Programming Framework The code generator maps • each description to an abstract class • each interaction contract to an abstract method By leveraging the GPL type checker, the framework • guides the implementation of what is required • forbids anything not specified in the design 42
A Programming Framework The code generator maps • each description to an abstract class • each interaction contract to an abstract method By leveraging the GPL type checker, the framework • guides the implementation of what is required • forbids anything not specified in the design different than what is proposed by ADLs or MDE 42
Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 43
Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 44
Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 45
Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } according to the interaction contract 46
Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 47
Implementing the Behavior context Presence as Boolean { source motion from MotionSensor; interaction { abstract class AbstractPresence { when provided motion from MotionSensor abstract boolean onMotionFromMotionSensor( boolean motion, Select select); get motion from MotionSensor } always publish } }  class Presence extends AbstractPresence { boolean onMotionFromMotionSensor( boolean motion, Select select) { return motion; } }  48
Implementing the Behavior when motion is detected ➡ there is presence when motion is not detected? ➡ ? The developer needs to ask all motion sensors class Presence extends AbstractPresence { boolean onMotionFromMotionSensor( boolean motion, Select select) { return motion; } }  49
Entity Selection Required when an entity is the interaction’s target Guide the developer with an embedded and type-safe DSL  50
Entity Selection entity MotionSensor { source motion as Boolean; attribute room as Integer; }  Select all motion sensors: select.motionSensors().all()  51
Entity Selection room = 1 room = 2 room = 3 1 2 3 entity MotionSensor { source motion as Boolean; attribute room as Integer; }  0 4 room = 0 room = 4 Select all motion sensors in rooms 1 to 3:  select.motionSensors().whereRoom(between(1,3)) 52
Commanding Entities entity Alarm { action OnOff { action OnOff; on(); } off(); }  room = 0 room = 4 Triggering all alarms: select.alarms().all().on();  53
Entity Selection Conformance entity Alarm { action OnOff { action OnOff; on(); } off(); }  room = 0 room = 4 Conformance It is not possible to send unsupported orders: select.alarms().all().start(); compile-time error 54
Entity Selection context Presence as Boolean { source motion from MotionSensor; interaction { when provided motion from MotionSensor get motion from MotionSensor always publish } }  Presence motion MotionSensor room = 0 room = 4 Conformance It is not possible to discover all kinds of entities: select.alarms().all(); compile-time error 55
Implementing the Behavior context Presence as Boolean { source motion from MotionSensor; interaction { abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor(...); when provided motion from MotionSensor } get motion from MotionSensor always publish } }  class Presence extends AbstractPresence { boolean onMotionFromMotionSensor( boolean motion, Select select) { if (motion) return true; MotionSensors sensors = select.motionSensors().all(); for (MotionSensor sensor : sensors) if (sensor.getMotion()) return true; return false; } }  56
Summary The developer is guided with • a support dedicated to the application  • an embedded DSL for entity selection Conformance is ensured by Conformance • generating a programming framework • leveraging a GPL type checker 57
Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 58
Evaluation of the Approach • Expressiveness • Usability • Productivity 59
Evaluation: Expressiveness Numerous domains • home-automation • avionics • graphical user interfaces • health-care • telecommunications • tier-system monitoring • etc. 60
Evaluation: Usability Context • 80 students during 3 years • sparse and oral-only documentation Results • 64 students completed the assignment • Identification of the interaction contracts 61
Evaluation: Productivity We measured the amount of code generated automatically 62
Evaluation: Productivity Specification 8% Implementation 10% Framework 82% 63
Evaluation: Productivity Specification 8% Implementation 10% Framework 82% ➡ 76% actually executed 63
Evaluation: Productivity Complexity of the developer’s code We used the Sonar platform to measure code quality through numerous metrics 64
Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code” McCabe cyclomatic complexity 64
Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code” McCabe cyclomatic complexity 1 3 7 10 ∞ 64
Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code” McCabe cyclomatic complexity “ quite well structured ” { 1 3 7 10 ∞ 64
Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code ” McCabe cyclomatic complexity “ quite well structured ” “ ” very poor quality { 1 3 7 10 ∞ 64
Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code ” McCabe cyclomatic complexity “ quite well structured ” “ ” very poor quality { 1 3 7 10 ∞ on average 64
Summary • The approach covers various domains • The frameworks are easy to use • Few code is required and this code is of good quality Pursuing this evaluation with software engineers 65
Results Scientific contributions   Conformance • A design language dedicated to SCC (ICSE’11) • The generation of a dedicated programming framework (GPCE’09) • The evaluation of this approach (submitted) Technical contributions • A compiler for the design language (9 KLoC) • A code generator targeting Java (4 KLoC) Dissemination • Demonstrations (PerCom’10), posters (SPLASH’10), visits (Bern, Potsdam) • Public release (http://diasuite.inria.fr) 66
A Research Vehicle This design language and code generator are part of a research project which involves • 4 industrial partnerships • 2 other research groups • > 20 real applications • 24/7 running platform • 28,000 lines of code 67
A Research Vehicle 7 PhD students leveraging DiaSpec and the generator • QoS (FASE’11) • error-handling (OOPSLA’10) • virtual testing (Mobiquitous’09 and ’10) • SIP (ICC’10, ICIN’09, IPTComm’08) • end-user programming (DSLWC’09) • security (ICPS’09) 68
Perspectives • Can we support other stages of the software life-cycle? requirements design code testing maintenance      • Can we transpose the approach to another paradigm? • Can we help creating such approaches? 69
70
Facilitating Evolution              • eases developer’s work by • showing mismatches    • leveraging development tools • ensures conformance all along the software life-cycle 71

More Related Content

PPT
What lies beneath the beautiful code?
Niranjan Sarade
 
PDF
A Generative Programming Approach to Developing Pervasive Computing Systems
Damien Cassou
 
PDF
The Parenscript Common Lisp to JavaScript compiler
Vladimir Sedach
 
PDF
parenscript-tutorial
tutorialsruby
 
PDF
Pharo tutorial at ECOOP 2013
Damien Cassou
 
ODP
Developing high-performance network servers in Lisp
Vladimir Sedach
 
KEY
Architecture-Driven Programming for Sense/Compute/Control Applications
Damien Cassou
 
PDF
Crash Course in Natural Language Processing (2016)
Vsevolod Dyomkin
 
What lies beneath the beautiful code?
Niranjan Sarade
 
A Generative Programming Approach to Developing Pervasive Computing Systems
Damien Cassou
 
The Parenscript Common Lisp to JavaScript compiler
Vladimir Sedach
 
parenscript-tutorial
tutorialsruby
 
Pharo tutorial at ECOOP 2013
Damien Cassou
 
Developing high-performance network servers in Lisp
Vladimir Sedach
 
Architecture-Driven Programming for Sense/Compute/Control Applications
Damien Cassou
 
Crash Course in Natural Language Processing (2016)
Vsevolod Dyomkin
 

Recently uploaded (20)

PDF
AI Unleashed - Shaping the Future -Starting Today - AIOUG Yatra 2025 - For Co...
Sandesh Rao
 
PPT
Coupa-Kickoff-Meeting-Template presentai
annapureddyn
 
PDF
Chapter 2 Digital Image Fundamentals.pdf
Getnet Tigabie Askale -(GM)
 
PPTX
Stamford - Community User Group Leaders_ Agentblazer Status, AI Sustainabilit...
Amol Dixit
 
PPTX
How to Build a Scalable Micro-Investing Platform in 2025 - A Founder’s Guide ...
Third Rock Techkno
 
PPTX
Coupa-Overview _Assumptions presentation
annapureddyn
 
PPTX
Smart Infrastructure and Automation through IoT Sensors
Rejig Digital
 
PDF
Revolutionize Operations with Intelligent IoT Monitoring and Control
Rejig Digital
 
PPTX
What-is-the-World-Wide-Web -- Introduction
tonifi9488
 
PDF
Software Development Company | KodekX
KodekX
 
PDF
Best ERP System for Manufacturing in India | Elite Mindz
Elite Mindz
 
PDF
SparkLabs Primer on Artificial Intelligence 2025
SparkLabs Group
 
PDF
Tea4chat - another LLM Project by Kerem Atam
a0m0rajab1
 
PDF
How-Cloud-Computing-Impacts-Businesses-in-2025-and-Beyond.pdf
Artjoker Software Development Company
 
PDF
Doc9.....................................
SofiaCollazos
 
PPTX
The-Ethical-Hackers-Imperative-Safeguarding-the-Digital-Frontier.pptx
sujalchauhan1305
 
PDF
Software Development Methodologies in 2025
KodekX
 
PDF
REPORT: Heating appliances market in Poland 2024
SPIUG
 
PDF
Automating ArcGIS Content Discovery with FME: A Real World Use Case
Safe Software
 
PDF
DevOps & Developer Experience Summer BBQ
AUGNYC
 
AI Unleashed - Shaping the Future -Starting Today - AIOUG Yatra 2025 - For Co...
Sandesh Rao
 
Coupa-Kickoff-Meeting-Template presentai
annapureddyn
 
Chapter 2 Digital Image Fundamentals.pdf
Getnet Tigabie Askale -(GM)
 
Stamford - Community User Group Leaders_ Agentblazer Status, AI Sustainabilit...
Amol Dixit
 
How to Build a Scalable Micro-Investing Platform in 2025 - A Founder’s Guide ...
Third Rock Techkno
 
Coupa-Overview _Assumptions presentation
annapureddyn
 
Smart Infrastructure and Automation through IoT Sensors
Rejig Digital
 
Revolutionize Operations with Intelligent IoT Monitoring and Control
Rejig Digital
 
What-is-the-World-Wide-Web -- Introduction
tonifi9488
 
Software Development Company | KodekX
KodekX
 
Best ERP System for Manufacturing in India | Elite Mindz
Elite Mindz
 
SparkLabs Primer on Artificial Intelligence 2025
SparkLabs Group
 
Tea4chat - another LLM Project by Kerem Atam
a0m0rajab1
 
How-Cloud-Computing-Impacts-Businesses-in-2025-and-Beyond.pdf
Artjoker Software Development Company
 
Doc9.....................................
SofiaCollazos
 
The-Ethical-Hackers-Imperative-Safeguarding-the-Digital-Frontier.pptx
sujalchauhan1305
 
Software Development Methodologies in 2025
KodekX
 
REPORT: Heating appliances market in Poland 2024
SPIUG
 
Automating ArcGIS Content Discovery with FME: A Real World Use Case
Safe Software
 
DevOps & Developer Experience Summer BBQ
AUGNYC
 
Ad

Featured (20)

PDF
2024 Trend Updates: What Really Works In SEO & Content Marketing
Search Engine Journal
 
PDF
Storytelling For The Web: Integrate Storytelling in your Design Process
Chiara Aliotta
 
PDF
Artificial Intelligence, Data and Competition – SCHREPEL – June 2024 OECD dis...
OECD Directorate for Financial and Enterprise Affairs
 
PDF
How to Leverage AI to Boost Employee Wellness - Lydia Di Francesco - SocialHR...
SocialHRCamp
 
PDF
2024 State of Marketing Report – by Hubspot
Marius Sescu
 
PDF
Everything You Need To Know About ChatGPT
Expeed Software
 
PDF
Product Design Trends in 2024 | Teenage Engineerings
Pixeldarts
 
PDF
How Race, Age and Gender Shape Attitudes Towards Mental Health
ThinkNow
 
PDF
AI Trends in Creative Operations 2024 by Artwork Flow.pdf
marketingartwork
 
PDF
Skeleton Culture Code
Skeleton Technologies
 
PDF
PEPSICO Presentation to CAGNY Conference Feb 2024
Neil Kimberley
 
PDF
Content Methodology: A Best Practices Report (Webinar)
contently
 
PPTX
How to Prepare For a Successful Job Search for 2024
Albert Qian
 
PDF
Social Media Marketing Trends 2024 // The Global Indie Insights
Kurio // The Social Media Age(ncy)
 
PDF
Trends In Paid Search: Navigating The Digital Landscape In 2024
Search Engine Journal
 
PDF
5 Public speaking tips from TED - Visualized summary
SpeakerHub
 
PDF
ChatGPT and the Future of Work - Clark Boyd
Clark Boyd
 
PDF
Getting into the tech field. what next
Tessa Mero
 
PDF
Google's Just Not That Into You: Understanding Core Updates & Search Intent
Lily Ray
 
PDF
How to have difficult conversations
Rajiv Jayarajah, MAppComm, ACC
 
2024 Trend Updates: What Really Works In SEO & Content Marketing
Search Engine Journal
 
Storytelling For The Web: Integrate Storytelling in your Design Process
Chiara Aliotta
 
Artificial Intelligence, Data and Competition – SCHREPEL – June 2024 OECD dis...
OECD Directorate for Financial and Enterprise Affairs
 
How to Leverage AI to Boost Employee Wellness - Lydia Di Francesco - SocialHR...
SocialHRCamp
 
2024 State of Marketing Report – by Hubspot
Marius Sescu
 
Everything You Need To Know About ChatGPT
Expeed Software
 
Product Design Trends in 2024 | Teenage Engineerings
Pixeldarts
 
How Race, Age and Gender Shape Attitudes Towards Mental Health
ThinkNow
 
AI Trends in Creative Operations 2024 by Artwork Flow.pdf
marketingartwork
 
Skeleton Culture Code
Skeleton Technologies
 
PEPSICO Presentation to CAGNY Conference Feb 2024
Neil Kimberley
 
Content Methodology: A Best Practices Report (Webinar)
contently
 
How to Prepare For a Successful Job Search for 2024
Albert Qian
 
Social Media Marketing Trends 2024 // The Global Indie Insights
Kurio // The Social Media Age(ncy)
 
Trends In Paid Search: Navigating The Digital Landscape In 2024
Search Engine Journal
 
5 Public speaking tips from TED - Visualized summary
SpeakerHub
 
ChatGPT and the Future of Work - Clark Boyd
Clark Boyd
 
Getting into the tech field. what next
Tessa Mero
 
Google's Just Not That Into You: Understanding Core Updates & Search Intent
Lily Ray
 
How to have difficult conversations
Rajiv Jayarajah, MAppComm, ACC
 
Ad

PhD thesis defense

  • 1. Leveraging Software Design to Guide the Development of Sense/Compute/Control Applications Damien Cassou
  • 2. Design is Crucial “The most important ingredient in ensuring a soft ware system’s long-term success is its design ” ICSE’11 c.f.p • A good design improves n1155: lf n162 ForLoop n536 start() n1704 n822 • collaboration ArcCond ArcSon ArcCond ArcCond n324 n1303 [name > len] n1304: Block Block error_mark visb: string = unk run_arch() ArcCond ArcCond ArcSon[2] [name <= len] stop_arch() n1702: Operator n1834: Operator • productivity op: string = fct_call ArcSon[1] op: string = fct_call n64: Literal ArcOpd[1] ArcOpd[2] value: string = 0 <<StartState . ActionState >> n7848 success << ActionState >> Suggest Itineraries NameRef n7848 n1966 RefersTo RefersTo NameRef ualFctParam error success n7840 RefersTo n7820: NameRef << EndState >> << EndState >> • maintenance NameRef value: string = len << ViewState >> startOver Display Itineraries RefersTo Try Again n471: Object success Cancel name: string = done n391: Function visb: string = unk name: string = exit static: bool = false << ActionState >> visb: string = pub extern: bool = false Select Itineraries virtual: string = non-virtual const: bool false assign book Instance RefersTo RefersTo n27: BuiltinType << SubFlowState >> finish << ActionState >> name: string = void n7828: NameRef n7846: NameRef Choose Itineraries Book Itineraries visb: string = unk name: string = done name: string = done success Confirmation 2
  • 3. Design Framework A good design requires a design framework which guides the architect with • a language n536 n1704 n822 start() n1155: lf n162 ForLoop ArcCond ArcSon ArcCond ArcCond n324 • a paradigm n1303 [name > len] n1304: Block Block error_mark visb: string = unk run_arch() ArcCond ArcCond ArcSon[2] [name <= len] stop_arch() n1702: Operator n1834: Operator op: string = fct_call ArcSon[1] op: string = fct_call n64: Literal ArcOpd[1] ArcOpd[2] value: string = 0 <<StartState . ActionState >> n7848 success << ActionState >> Suggest Itineraries NameRef n7848 n1966 RefersTo RefersTo NameRef ualFctParam error success n7840 RefersTo n7820: NameRef << EndState >> << EndState >> NameRef value: string = len << ViewState >> startOver Display Itineraries RefersTo Try Again n471: Object success Cancel name: string = done n391: Function visb: string = unk name: string = exit static: bool = false << ActionState >> visb: string = pub extern: bool = false Select Itineraries virtual: string = non-virtual const: bool false assign book Instance RefersTo RefersTo n27: BuiltinType << SubFlowState >> finish << ActionState >> name: string = void n7828: NameRef n7846: NameRef Choose Itineraries Book Itineraries visb: string = unk name: string = done name: string = done success Confirmation 3
  • 4. Programming Framework A good implementation requires a programming framework which guides the developer with • abstractions • services 4
  • 5. Conformance An implementation must conform to its design n536 n1704 n822 start() n1155: lf n162 ForLoop ArcCond ArcSon ArcCond ArcCond n324 n1303 [name > len] n1304: Block Block error_mark visb: string = unk run_arch() ArcCond ArcCond ArcSon[2] [name <= len] stop_arch() n1702: Operator n1834: Operator op: string = fct_call ArcSon[1] op: string = fct_call n64: Literal ArcOpd[1] ArcOpd[2] value: string = 0 <<StartState . ActionState >> n7848 success << ActionState >> Suggest Itineraries NameRef n7848 n1966 RefersTo RefersTo NameRef ualFctParam error success n7840 RefersTo n7820: NameRef << EndState >> << EndState >> NameRef value: string = len << ViewState >> startOver Display Itineraries RefersTo Try Again n471: Object success Cancel name: string = done n391: Function visb: string = unk name: string = exit static: bool = false << ActionState >> visb: string = pub extern: bool = false Select Itineraries virtual: string = non-virtual const: bool false assign book Instance RefersTo RefersTo n27: BuiltinType << SubFlowState >> finish << ActionState >> name: string = void n7828: NameRef n7846: NameRef Choose Itineraries Book Itineraries visb: string = unk name: string = done name: string = done success Confirmation 5
  • 6. Requirements 1. A design framework to guide the architect  2. A programming framework to guide the developer  3. A guaranteed conformance of the implementation relatively to the design Conformance 6
  • 7. Related Works Design  Implementation  Conformance Conformance 7
  • 8. Related Works GPL Design  Implementation +  Conformance Conformance Java 8
  • 9. Related Works GPL Library Design  Implementation + ++  Conformance Conformance Spring 9
  • 10. Related Works GPL Library ADL Design ++  Implementation + ++  Conformance Conformance C2 10
  • 11. Related Works GPL Library ADL ADL++ Design ++ +  Implementation + ++ +  Conformance + Conformance ArchJava 11
  • 12. Thesis A paradigm-oriented framework for both design and implementation which maintains conformance all along the software life-cycle 12
  • 13. The Paradigm Sense/Compute/Control (SCC) “applications that interact with an environment” Environment 13
  • 14. The SCC Paradigm sense compute Environment control 14
  • 15. The SCC Paradigm sense GPS, flight plan compute compute direction control control aileron, engine 15
  • 16. The SCC Paradigm motion detection sense compute Environment intrusion? control trigger alarms 16
  • 17. The SCC Paradigm Covers various domains • pervasive computing • tier-system monitoring Environment • avionics • robotics • ... 17
  • 18. Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 18
  • 19. Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 19
  • 20. Design Language Environment orders 20
  • 21. Design Language sources Environment • sources sense the environment orders 20
  • 22. Design Language sources Environment actions • sources sense the environment • actions impact the environment orders 20
  • 23. Design Language a concept for handling the interaction with the environment sources Environment actions • sources sense the environment • actions impact the environment orders 20
  • 24. Design Language entity Keypad { source keycode as Integer; a concept for handling action UpdateSt;  } the interaction with the environment action UpdateSt { updateStatus(message as String); } sources Entities Environment actions • sources sense the environment • actions impact the environment orders 20
  • 25. Design Language entity Keypad { source keycode as Integer; a concept for handling action UpdateSt;  } the interaction with the environment action UpdateSt { updateStatus(message as String); } sources Entities Environment actions 1 entity description for potentially many implementations orders 21
  • 26. Design Language entity Keypad { source keycode as Integer; a concept for handling action UpdateSt;  } the interaction with the environment action UpdateSt { updateStatus(message as String); } sources Entities Environment actions 1 entity description for potentially many instances orders 22
  • 27. Design Language entity Keypad { source keycode as Integer; action UpdateSt;  attribute room as Integer; } action UpdateSt { updateStatus(message as String); } sources Entities Environment actions 1 entity description for potentially many instances ➡ attributes to characterize instances orders (color, location, reliability, etc.) 23
  • 28. Design Language raw data a concept to refine raw data sources Entities Environment actions orders 24
  • 29. Design Language raw data a concept to refine raw data context operators sources refined information Entities Environment actions context BuildingSecured as Boolean { source keycode from Keypad; } orders  24
  • 30. Design Language raw data a concept to take decisions based on application-level data context operators sources refined information Entities Environment actions orders 25
  • 31. Design Language raw data a concept to take decisions based on application-level data context operators sources refined information Entities Environment actions control operators controller BuildingManager { context BuildingSecured; orders orders action UpdateSt on Keypad; }  25
  • 32. Design Language raw data The language features concepts Compute dedicated to the SCC paradigm context Sense operators sources refined information Entities Environment actions control operators orders orders Control 26
  • 33. Application logic Environment handling context operators sources refined information Entities Environment actions control operators orders 27
  • 34. context operators Information creation refined information control operators Information use orders 28
  • 35. Design Language raw data context operators sources refined information Entities Environment actions control operators orders orders 29
  • 36. Design Language actions control operators context operators orders sources 30
  • 37. Design Language actions control operators context operators orders sources 31
  • 38. Case Study: Anti-Intrusion actions control operators context operators sources 32
  • 39. Case Study: Anti-Intrusion actions control operators Intrusion context operators sources 32
  • 40. Case Study: Anti-Intrusion Alarm actions OnOff Intrusion control Manager operators Intrusion context operators sources 32
  • 41. Case Study: Anti-Intrusion Alarm actions OnOff Intrusion control Manager operators Intrusion context operators Building Presence Secured sources 32
  • 42. Case Study: Anti-Intrusion Alarm actions OnOff Intrusion control Manager operators Intrusion context operators Building Presence Secured keycode motion Keypad MotionSensor sources 32
  • 43. Case Study: Anti-Intrusion Keypad Alarm actions UpdateSt OnOff Security Intrusion control Manager Manager operators Intrusion context operators Building Presence Secured keycode motion Keypad MotionSensor sources 32
  • 44. Case Study: Anti-Intrusion Keypad Alarm Mailer actions UpdateSt OnOff Send Security Intrusion control Manager Manager operators Intrusion context operators Building Scene Presence Secured Image keycode motion image Keypad MotionSensor Camera sources 32
  • 45. Case Study: Anti-Intrusion Keypad Alarm Mailer actions UpdateSt OnOff Send Security Intrusion control Manager Manager operators Intrusion context operators Building Scene Presence Secured Image keycode motion image Keypad MotionSensor Camera sources 33
  • 46. Case Study: Anti-Intrusion Intrusion Building Presence Secured 34
  • 47. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured what the architect has in mind 34
  • 48. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured what the architect has in mind 34
  • 49. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured what the architect has in mind 34
  • 50. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations 35
  • 51. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
  • 52. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
  • 53. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
  • 54. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Building Presence Secured 35
  • 55. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured 35
  • 56. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured 35
  • 57. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured 35
  • 58. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured the architect should express what he has in mind 35
  • 59. Case Study: Anti-Intrusion Intrusion Intrusion Building Building Presence Presence Secured Secured possible interpretations Intrusion Intrusion Building Building Presence Presence Secured Secured the architect should express what he has in mind  ➡interaction contracts 35
  • 60. Interaction Contracts context operator 36
  • 61. Interaction Contracts request 1 Activation condition 1 context operator 36
  • 62. Interaction Contracts 1 Activation condition context operator event 1 36
  • 63. Interaction Contracts 1 Activation condition 2 Data requirement context operator event 1 2 request 2 request source context Entity operator 36
  • 64. Interaction Contracts 1 Activation condition 3 event 2 Data requirement context operator 3 Emission event 1 2 request 2 request source context Entity operator 36
  • 65. Interaction Contracts 1 Activation condition 2 Data requirement Intrusion 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { when provided Presence get BuildingSecured maybe publish } } 37 
  • 66. Interaction Contracts 1 Activation condition 2 Data requirement Intrusion 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence get BuildingSecured maybe publish } } 37 
  • 67. Interaction Contracts 1 Activation condition 2 Data requirement Intrusion 2 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence 2 get BuildingSecured maybe publish } } 37 
  • 68. Interaction Contracts 1 Activation condition 3 2 Data requirement Intrusion 2 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence 2 get BuildingSecured 3 maybe publish } } 37 
  • 69. Interaction Contracts 1 Activation condition 3 2 Data requirement Intrusion 2 1 3 Emission Building context Intrusion as Boolean { Presence context Presence; Secured context BuildingSecured; interaction { 1 when provided Presence 2 get BuildingSecured 3 maybe publish related to automata } approaches } 37 
  • 70. Summary  A design framework consisting of a design language, DiaSpec, which guides the architect by offering • concepts dedicated to the SCC paradigm • a separation between environment handling and logic • a separation between information creation and use • a dedicated description of interactions 38
  • 71. Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 39
  • 72. A Programming Framework Generated from the Design DiaSpec GPL         40
  • 73. A Programming Framework Generated from the Design DiaSpec GPL         • separates 2 different roles with 2 different languages • leverages GPL tools, libraries and expertise • ensures conformance automatically 40
  • 74. A Programming Framework how to make a programming framework conform to a particular design? 41
  • 75. A Programming Framework The code generator maps • each description to an abstract class • each interaction contract to an abstract method 42
  • 76. A Programming Framework The code generator maps • each description to an abstract class • each interaction contract to an abstract method By leveraging the GPL type checker, the framework • guides the implementation of what is required • forbids anything not specified in the design 42
  • 77. A Programming Framework The code generator maps • each description to an abstract class • each interaction contract to an abstract method By leveraging the GPL type checker, the framework • guides the implementation of what is required • forbids anything not specified in the design different than what is proposed by ADLs or MDE 42
  • 78. Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 43
  • 79. Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 44
  • 80. Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 45
  • 81. Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } according to the interaction contract 46
  • 82. Generation context Presence as Boolean { source motion from MotionSensor; interaction { Presence when provided motion from MotionSensor get motion from MotionSensor motion always publish MotionSensor } }  abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor( boolean motion, Select select); } 47
  • 83. Implementing the Behavior context Presence as Boolean { source motion from MotionSensor; interaction { abstract class AbstractPresence { when provided motion from MotionSensor abstract boolean onMotionFromMotionSensor( boolean motion, Select select); get motion from MotionSensor } always publish } }  class Presence extends AbstractPresence { boolean onMotionFromMotionSensor( boolean motion, Select select) { return motion; } }  48
  • 84. Implementing the Behavior when motion is detected ➡ there is presence when motion is not detected? ➡ ? The developer needs to ask all motion sensors class Presence extends AbstractPresence { boolean onMotionFromMotionSensor( boolean motion, Select select) { return motion; } }  49
  • 85. Entity Selection Required when an entity is the interaction’s target Guide the developer with an embedded and type-safe DSL  50
  • 86. Entity Selection entity MotionSensor { source motion as Boolean; attribute room as Integer; }  Select all motion sensors: select.motionSensors().all()  51
  • 87. Entity Selection room = 1 room = 2 room = 3 1 2 3 entity MotionSensor { source motion as Boolean; attribute room as Integer; }  0 4 room = 0 room = 4 Select all motion sensors in rooms 1 to 3:  select.motionSensors().whereRoom(between(1,3)) 52
  • 88. Commanding Entities entity Alarm { action OnOff { action OnOff; on(); } off(); }  room = 0 room = 4 Triggering all alarms: select.alarms().all().on();  53
  • 89. Entity Selection Conformance entity Alarm { action OnOff { action OnOff; on(); } off(); }  room = 0 room = 4 Conformance It is not possible to send unsupported orders: select.alarms().all().start(); compile-time error 54
  • 90. Entity Selection context Presence as Boolean { source motion from MotionSensor; interaction { when provided motion from MotionSensor get motion from MotionSensor always publish } }  Presence motion MotionSensor room = 0 room = 4 Conformance It is not possible to discover all kinds of entities: select.alarms().all(); compile-time error 55
  • 91. Implementing the Behavior context Presence as Boolean { source motion from MotionSensor; interaction { abstract class AbstractPresence { abstract boolean onMotionFromMotionSensor(...); when provided motion from MotionSensor } get motion from MotionSensor always publish } }  class Presence extends AbstractPresence { boolean onMotionFromMotionSensor( boolean motion, Select select) { if (motion) return true; MotionSensors sensors = select.motionSensors().all(); for (MotionSensor sensor : sensors) if (sensor.getMotion()) return true; return false; } }  56
  • 92. Summary The developer is guided with • a support dedicated to the application  • an embedded DSL for entity selection Conformance is ensured by Conformance • generating a programming framework • leveraging a GPL type checker 57
  • 93. Contributions 1. A paradigm-specific design framework 2. A programming framework dedicated to a design 3. An evaluation of the approach 58
  • 94. Evaluation of the Approach • Expressiveness • Usability • Productivity 59
  • 95. Evaluation: Expressiveness Numerous domains • home-automation • avionics • graphical user interfaces • health-care • telecommunications • tier-system monitoring • etc. 60
  • 96. Evaluation: Usability Context • 80 students during 3 years • sparse and oral-only documentation Results • 64 students completed the assignment • Identification of the interaction contracts 61
  • 97. Evaluation: Productivity We measured the amount of code generated automatically 62
  • 98. Evaluation: Productivity Specification 8% Implementation 10% Framework 82% 63
  • 99. Evaluation: Productivity Specification 8% Implementation 10% Framework 82% ➡ 76% actually executed 63
  • 100. Evaluation: Productivity Complexity of the developer’s code We used the Sonar platform to measure code quality through numerous metrics 64
  • 101. Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code” McCabe cyclomatic complexity 64
  • 102. Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code” McCabe cyclomatic complexity 1 3 7 10 ∞ 64
  • 103. Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code” McCabe cyclomatic complexity “ quite well structured ” { 1 3 7 10 ∞ 64
  • 104. Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code ” McCabe cyclomatic complexity “ quite well structured ” “ ” very poor quality { 1 3 7 10 ∞ 64
  • 105. Evaluation: Productivity Complexity of the developer’s code “ number of linearly independent paths in a source code ” McCabe cyclomatic complexity “ quite well structured ” “ ” very poor quality { 1 3 7 10 ∞ on average 64
  • 106. Summary • The approach covers various domains • The frameworks are easy to use • Few code is required and this code is of good quality Pursuing this evaluation with software engineers 65
  • 107. Results Scientific contributions   Conformance • A design language dedicated to SCC (ICSE’11) • The generation of a dedicated programming framework (GPCE’09) • The evaluation of this approach (submitted) Technical contributions • A compiler for the design language (9 KLoC) • A code generator targeting Java (4 KLoC) Dissemination • Demonstrations (PerCom’10), posters (SPLASH’10), visits (Bern, Potsdam) • Public release (http://diasuite.inria.fr) 66
  • 108. A Research Vehicle This design language and code generator are part of a research project which involves • 4 industrial partnerships • 2 other research groups • > 20 real applications • 24/7 running platform • 28,000 lines of code 67
  • 109. A Research Vehicle 7 PhD students leveraging DiaSpec and the generator • QoS (FASE’11) • error-handling (OOPSLA’10) • virtual testing (Mobiquitous’09 and ’10) • SIP (ICC’10, ICIN’09, IPTComm’08) • end-user programming (DSLWC’09) • security (ICPS’09) 68
  • 110. Perspectives • Can we support other stages of the software life-cycle? requirements design code testing maintenance      • Can we transpose the approach to another paradigm? • Can we help creating such approaches? 69
  • 111. 70
  • 112. Facilitating Evolution              • eases developer’s work by • showing mismatches    • leveraging development tools • ensures conformance all along the software life-cycle 71

Editor's Notes