Model-based Variability Management (tutorial at ECOOP/ECMFA/ECSA)

Model-based Variability Management (tutorial at ECOOP/ECMFA/ECSA)

The customization of almost everything is observed in a wide range of domains. Many organizations should address the challenge of extending, changing, customizing or configuring numerous kinds of systems and artefacts (requirements, components, services, languages, architectural or design models, codes, user interfaces, etc.) for use in a particular context. As a result, modeling and managing variability of such systems and artefacts is a crucial activity in a growing number of software engineering contexts (e.g., software product lines, dynamic adaptive architectures). Numerous model-based techniques have been proposed and usually consist in i) a variability model (e.g., a feature model), ii) a model (e.g., a class diagram) expressed in a domain-specific modeling language (e.g., Unified Modelling language), and iii) a realization layer that maps and transforms variation points into model elements. Based on a selection of desired features in the variability model, a derivation engine can automatically synthesise customized models – each model corresponding to an individual product. In this tutorial, we present the foundations and tool-supported techniques of state-of-the-art variability modeling technologies. In the first part, we briefly exemplify the management of variability in some systems/artefacts (design models, languages, product configurators). We introduce the Common Variability Language (CVL), a representative approach and ongoing effort involving both academic and industry partners to promote standardization variability modeling technology. In the second part, we focus on feature models the most popular notation to formally represent and reason about commonality and variability of a software system. Feature modelling languages and tools, directly applicable to a wide range of model-based variability problems and application domains, are presented. The FAMILIAR language and environment is used to perform numerous management operations like the import, export, compose, decompose, edit, configure, compute diffs, refactor, reverse engineer, test, or reason about (multiple) feature models. We describe their theoretical foundations, efficient implementations, and how these operations can be combined to realize complex variability management tasks. In the third part, we show how to combine feature models and other modeling artefacts. We revisit the examples given in the first part of the tutorial, using the Kermeta workbench and familiarCVL, an implementation of CVL. Finally we present some of the ongoing challenges for variability modeling. At the end of the tutorial, participants (being practitioners or academics, beginners or advanced) will learn languages, tools and novel variability modeling techniques they can directly use in their industrial contexts or as part of their research.

Bb59170948068df05f09ab11a9f3f35e?s=128

FAMILIAR project

July 03, 2013
Tweet

Transcript

  1. Mathieu  Acher,  Benoit  Combemale,  Olivier  Barais   Model-­‐Based    

    Variability  Management    
  2. 2   Research  in  so6ware   engineering.   -­‐  8

     faculty  members   -­‐  35  researchers  and   engineers  on  projects  
  3. We’re  hiring!     engineers,  PhD  students,  post-­‐docs   Variability

     /  Product  lines     Model-­‐driven  Engineering   Language  Engineering  (e.g.,  DSLs)   Scala   3   3   European  Projects       Industrial  CollaboraCons     Academics  partners  
  4. Acknowledgments  (la  famille)   Marianela  Ciolfi  Felice     Joao

     Bosco  Ferreira  Filho   Guillaume  Bécan   Suresh  Pilay     Sana  Ben  Nasr   (MSc/PhD  students,     University  of  Rennes  1)     Prof.  Philippe  Collet   Prof.  Philippe  Lahire     (University  of  Nice  Sophia  AnWpolis)     Prof.  Robert  B.  France     (Colorado  State  University)     Prof.  Patrick  Heymans     (University  of  Namur)  
  5. Audience   •  No  pre-­‐requisite  background!   •  Targeted  Audience

      •  Academics  or  pracWWoners     •  Curious  guys:  e.g.,  PhD  students  or  modellers  unaware  of…     –  Variability  and  so6ware  product  lines  (SPLs)   –  Variability  modelling     –  ConfiguraWon   •  MDE  guys:  people  involved  or  interested  in  the  development  of   model  management  tools   –  e.g.,  model  composiWon/decomposiWon   •  SPL  guys:  advances  that  want  to  learn  new  techniques   5  
  6. At  the  end  of  the  tutorial…   •  You  will

     have  an  overview  of  what’s  going  on  in  the  field  of     variability  and  model-­‐based  so6ware  product  line  engineering   •  You  will  be  able  to  go  further  with  the  languages  and  modelling   techniques   •  so  to  reuse  them  in  pracWcal  or  academic  contexts     •  SupporWng  material:   hbps://github.com/FAMILIAR-­‐project/familiar-­‐documentaWon/blob/ master/presentaWons/EC2013/README.md     •  slides  of  the  tutorial   •  related  arWcles,     •  FAMILIAR  scripts,   •  CVL  models,   •  and  packaged  tools  to  interacWvely  play  with  the  models  during  the   tutorial   6  
  7. Differences  with  previous  tutorials  at   SPLC’12  /  MODELS’12  

    •  Larger  perspecWve/moWvaWon   •  Including  modelling/language/architectural  examples     •  Not  only  about  feature  models   •  not  only  about  FAMILIAR   •  but  new  techniques  for  reverse  engineering  (VaMoS’13)  and  composing   (MODELS’13)  feature  models  will  be  presented     •  Model-­‐based  product  line  engineering   •  Common  Variability  Language  (CVL)   FAMILIAR  is  now  a  project    not  only  a  language  for  managing  feature  models!   7  
  8. [MOTIVATION/PROBLEM]  Why  modeling  and  managing  Variability   does  and  will

     maber  (30’)   [SOLUTION  FOR  MANAGING  FEATURE  MODELS]  Managing  Variability   Models  with  FAMILIAR  (1h45’)       [SOLUTION  FOR  MODEL-­‐BASED  DERIVATION  OF  PRODUCT]  Model-­‐based   variability  engineering:  examples,  support  and  open  issues   (45’)   8   Plan  
  9. [MOTIVATION/PROBLEM]  Why  modeling  and  managing  Variability   does  and  will

     maber  (30’)   [SOLUTION  FOR  MANAGING  FEATURE  MODELS]  Managing  Variability   Models  with  FAMILIAR  (1h45’)       [SOLUTION  FOR  MODEL-­‐BASED  DERIVATION  OF  PRODUCT]  Model-­‐based   variability  engineering:  examples,  support  and  open  issues   (45’)   9   Plan  
  10. 10   So6ware-­‐intensive  systems   come  in  many  variants  

     
  11. 11  

  12. Linux   Kernel  

  13. Database   Engine  

  14. Printer   Firmware  

  15. Features  in  MicrosoS  Office   15  

  16. None
  17. 17  

  18. Variability     “the  ability  of  a  system  to  be

     efficiently  extended,   changed,  customized  or  configured  for  use  in  a   parCcular  context”     Mikael  Svahnberg,  Jilles  van  Gurp,  and  Jan  Bosch  (2005)  
  19. «  A  set  of  programs  is  considered  to  consWtute  

    a  family,  whenever  it  is  worthwhile  to  study   programs  from  the  set  by  first  studying  the   common  properCes  of  the  set  and  then   determining  the  special  properCes  of  the   individual  family  members  »             David  L.  Parnas  —  ‘‘On  the  design  and  development  of  program   families’’  in  TransacCons  on  SoSware  Engineering,  SE-­‐2(1):1–9,  1976     19   aka  Variability  
  20. Variability     “the  ability  of  a  system  to  be

     efficiently   extended,  changed,  customized  or   configured  for  use  in  a  parCcular  context”     Mikael  Svahnberg,  Jilles  van  Gurp,  and  Jan  Bosch  (2005)       20  
  21. 21   21   Extensible  architectures   (eg  plugins-­‐based)  

    ConfiguraCon   files   System   Preferences   Configurators   Source  code   Build   systems   Comparison  of  *   Structural  or  behavorial     models   External  Variability   Internal  Variability   Variability  @  run.Cme  
  22. 22   «  Feature  Model  ExtracWon  from  Large  CollecWons  of

     Informal  Product  DescripWons  »     Jean-­‐Marc  Davril,  Edouard  Delfosse,  Negar  Hariri,  Mathieu  Acher,  Jane  Cleland-­‐Huang,  Patrick   Heymans  (ESEC/FSE’13)  
  23. 23   «  The  Anatomy  of  a  Sales  Configurator:  An

     Empirical  Study  of  111  Cases  »  Ebrahim  Khalil  Abbasi,   Arnaud  Hubaux,  Mathieu  Acher,  QuenWn  Boucher,  and  Patrick  Heymans  (CAiSE’13)  
  24. 24   «  ExtracWon  and  EvoluWon  of  Architectural  Variability  Models

     in  Plugin-­‐based  Systems  »       Mathieu  Acher,  Anthony  Cleve,  Philippe  Collet,  Philippe  Merle,  Laurence  Duchien,  Philippe   Lahire  ECSA/SoSyM’13  
  25. If  you’re  able  to  master  variability…   •  Reduce  development

     costs     •  Reduce  cerWficaWon  costs     •  Shorten  Wme-­‐to-­‐market     •  But,  are  you  able?     – developing,  verifying,  cerWfying  billions  of  variants  is   challenging!     25  
  26. Variability = Complexity ChrisWan  Kästner  slide  

  27. a  unique  variant  for  every   person  on  this  planet

      33  features   opWonal,  independent   ChrisWan  Kästner  slide  
  28. 320  features     more  variants  than  esWmated    

     atoms  in  the  universe   opWonal,  independent  
  29. 2000  features   10000   features   ChrisWan  Kästner  slide

     
  30. 30       Avoid  solving  the  same  problem!  

         2,  3…n  Cmes     AutomaCon?  
  31. 31   Unused  flexibility  

  32. 32   Illegal  variant  

  33.     Goal:  So6ware  mass  customizaWon     /  AdapWve

     and  configurable  systems     Problem:  Variability  =  Complexity     Approach:  Model-­‐based  variability  management   33   Why  managing  Variability     does  (and  will)  maier  
  34. 34   So6ware-­‐intensive  systems   come  in  many  variants  

      Model-­‐based     Variability  Management  
  35.   Modeling  Variability     CommunicaCve     AnalyCc  

      GeneraCve     35  
  36. 36  

  37. None
  38. 38   Factoring  out  commonaliCes    for  Reuse  [Krueger  et

     al.,  1992]  [Jacobson  et  al.,  1997]               Managing  variabiliCes      for  So6ware  Mass  CustomizaCon  [Bass  et  al.,  1998]  [Krueger  et  al.,  2001],  [Pohl  et  al.,  2005]      
  39. Mobile 3G+ 3G GPS Maps Camera ü   ü  

    ü   Mobile 3G+ 3G GPS Maps Camera Domain/Variability  Model   ConfiguraCon   SoSware  Generator   Domain  Artefacts       Domain     Engineering   ApplicaCon     Engineering   «  the  investments  required  to  develop  the  reusable  arBfacts  during   domain  engineering,  are  outweighed  by  the  benefits  of  deriving  the   individual  products  during  applica.on  engineering  »   Jan  Bosch  et  al.  (2004)      
  40. 40   99%  domain  engineering,     1%  applicaCon  engineering?

      –  specifies  what  you  want  (click,  click,  click)  a  customized   product  is  automaWcally  built  for  you   –  Iterate  the  process  for  n  products   Amount of effort Application Engineering More Sophisticated Technology Domain Engineering
  41. Variability  AbstracCon   Model  (VAM)   ConfiguraCon   (resoluCon  model)

      Domain  Artefacts   (e.g.,  models)   SoSware  Generator   (derivaCon  engine)   ü   ü   Variability   RealizaCon   Model   (VRM)  
  42. 42  

  43. Configurations Derivation Process Models of the “system” Feature Model How

    to realize the variability
  44. (another  research  area/applicaWon:   adapWve  systems  aka  dynamic  so6ware  product

     lines   Models@run.Wme)   hip://www.kevoree.org   from  Cloud  stack  to   embedded  devices  
  45. None
  46. Variability  Handling  in  AUTOSAR   Body   control   Low-­‐end

     light-­‐ control   AdapWve-­‐curve   light-­‐control   Feature  Modeling   (Variability  abstracWon)   Generic  Template   (Variability  RealizaWon)   LightType   …   System   constant   Low  End   High  End   Car   1..1 Feature   v.  4.04  (upcoming)   v.  4.03   Low  End  ==  1   High  End  ==  1   VariaWon   point   Adapted  from  the  CVL  tutorial  at  SPLC’12  by  Oystein  Haugen,  Andrezj  Wasowski,  Krzysztof  Czarnecki    
  47. Variability  Handling  in  AUTOSAR  (2)   Feature  Modeling   (Variability

     abstracWon)   Generic  Template   (Variability  RealizaWon)   Low  End   High  End   Car   1..1 Body   control   Low-­‐end   light-­‐control  
  48. VariaCon  Point  Types   •  Variability  is  applied  to  different

     parts  of  the   metamodel   – AggregaWon,  associaWon,  abribute  value,  property   set   •  ResulWng  variability   – OpWonal  component   – OpWonal  port   – OpWonal  connector   – Parameter  variability   Component   Port   Adapted  from  the  CVL  tutorial  at  SPLC’12  by  Oystein  Haugen,  Andrezj  Wasowski,  Krzysztof  Czarnecki    
  49. 49   «  Mapping  Features  to  Models:  A  Template  Approach

     Based  on  Superimposed  Variants»   Krzysztof  Czarnecki  and  Michal  Antkiewicz  GPCE’05  
  50. 50   «  Mapping  Features  to  Models:  A  Template  Approach

     Based  on  Superimposed  Variants»   Krzysztof  Czarnecki  and  Michal  Antkiewicz  GPCE’05  
  51. 51   «  Mapping  Features  to  Models:  A  Template  Approach

     Based  on  Superimposed  Variants»   Krzysztof  Czarnecki  and  Michal  Antkiewicz  GPCE’05  
  52. 52   «  Mapping  Features  to  Models:  A  Template  Approach

     Based  on  Superimposed  Variants»   Krzysztof  Czarnecki  and  Michal  Antkiewicz  GPCE’05  
  53. Safe  composiWon?  No!   53   «Verifying  Feature-­‐Based  Model  Templates

     Against   Well-­‐Formedness  OCL  Constraints  »  Krzysztof  Czarnecki  Krzysztof  Pietroszek  GPCE’06  
  54. Ooops   54   «Verifying  Feature-­‐Based  Model  Templates  Against  

    Well-­‐Formedness  OCL  Constraints  »  Krzysztof  Czarnecki  Krzysztof  Pietroszek  GPCE’06  
  55. Another  approach   55   «  Reconciling  AutomaWon  and  Flexibility

     in  Product  DerivaWon  »  Gilles  Perrouin,  Jacques  Klein,   Nicolas  Guelfi,  Jean-­‐Marc  Jézéquel  SPLC’08  
  56. 56   «  Reconciling  AutomaWon  and  Flexibility  in  Product  DerivaWon

     »  Gilles  Perrouin,  Jacques  Klein,   Nicolas  Guelfi,  Jean-­‐Marc  Jézéquel  SPLC’08   Merging-­‐based  DerivaCon  of  Product  
  57. None
  58. Variability  at  the  language  level     58 Variability  in

      Metamodeling   •  SemanWc  variaWon  point   •  DSML  Families   •  Knowledge  capitalizaWon   •  Language  Engineering     Variability  in   Modeling   Variability Variability
  59.          

         Engineering  SemanCcs  in  Modeling  Languages   59 Abstract Syntax (AS) Concrete Syntax (CS) Semantics Domain (SD) Mac Mas •  Variability  in  metamodeling  (DSML  families,  variaWon  point...):   –  Abstract  syntax:  staWc  introducWon  (AOM),  inheritance  (OOP)   –  Concrete  syntax:  view  point  (OBEO  Designer)   –  SemanWcs:  sWll  a  problem!  how  to  define  and  manage  semanBc   variability  (in  the  DSML  and  the  associated  tools)?  
  60. DSL4   DSL3   DSL2   DSL1   Language  Family

      (expresiveness,  semanWc  variaWon  point,     implementaWon  variaWon  point,  viewpoints,  tooling,  etc.)   RM   dsl1   RM   dsl2   RM   dsl3   RM   dsl4   Challenge1:  Modular   Language  Design   Challenge3:   Language   ComposiWon   Challenge2:   Variability  Modeling   «  Variability  Management  in  Modeling  Languages  »  Suresh  Pilay  PhD  thesis  (ongoing)  
  61. DSL   Variability   model   CVL   Base  

      model   Generic  &     Standardized   resoluCon   models   Focused  on     a  domain   Execute  CVL       Resolved     models   DescripWon  of   possible   variaWons  in   the  system   Domain   model  of  a   parWcular   family  of   system   SelecWon  of  a  set   of  opWons  in  the   variaWon  model   Family  of  systems  fully   described  in  the   domain  specific   language.   All  regular  DSL  tools   can  be  applied  to  these   models   61   RealizaCon   model   Language Units Language Features how to realize the features Configuration of languages Derivation Process Languages «  Variability  Management  Modeling  Languages  »  Suresh  Pilay  PhD  thesis  (ongoing)  
  62. None
  63. 63   «  ExtracWon  and  EvoluWon  of  Architectural  Variability  Models

     in  Plugin-­‐based  Systems  »       Mathieu  Acher,  Anthony  Cleve,  Philippe  Collet,  Philippe  Merle,  Laurence  Duchien,  Philippe   Lahire  ECSA/SoSyM’13  
  64. 64   «  ExtracWon  and  EvoluWon  of  Architectural  Variability  Models

     in  Plugin-­‐based  Systems  »       Mathieu  Acher,  Anthony  Cleve,  Philippe  Collet,  Philippe  Merle,  Laurence  Duchien,  Philippe   Lahire  ECSA/SoSyM’13   FraSCAti SCAParser Java Compiler JDK6 JDT Optional Mandatory Alternative- Group Or-Group Assembly Factory rest http Binding MMFrascati Component Factory Metamodel MMTuscany constraints rest requires MMFrascati http requires MMTuscany FM1 Variability  Model  
  65. 65   Variability  Model   FraSCAti SCAParser Java Compiler JDK6

    JDT Optional Mandatory Alternative- Group Or-Group Assembly Factory rest http Binding MMFrascati Component Factory Metamodel MMTuscany constraints rest requires MMFrascati http requires MMTuscany FM1 FraSCAC  Architecture   Set  of    Safe   Variants   authorized  by   FraSCAC   Scope  is   too  large  
  66. Illegal    Variant     66  

  67. 67   FraSCAC  Architecture   FraSCAti SCAParser Java Compiler JDK6

    JDT Optional Mandatory Alternative- Group Or-Group Assembly Factory rest http Binding MMFrascati Component Factory Metamodel MMTuscany constraints rest requires MMFrascati http requires MMTuscany FM1 Feature  Model   FraSCAti SCAParser Java Compiler JDK6 JDT Optional Mandatory Alternative- Group Or-Group Assembly Factory rest http Binding MMFrascati Component Factory Metamodel MMTuscany constraints rest requires MMFrascati http requires MMTuscany FM1 ConfiguraCon   Derived  FraSCAC  Architecture  
  68. [MOTIVATION/PROBLEM]  Why  modeling  and  managing  Variability   does  and  will

     maber  (30’)   [SOLUTION  FOR  MANAGING  FEATURE  MODELS]  Managing  Variability   Models  with  FAMILIAR  (1h45’)       [SOLUTION  FOR  MODEL-­‐BASED  DERIVATION  OF  PRODUCT]  Model-­‐based   variability  engineering:  examples,  support  and  open  issues   (45’)   68   Plan  
  69. Variability  Model   ConfiguraCon   Domain  Artefacts  (e.g.,  source  code)

      SoSware  Generator   Modeling   variability     is  crucial   ü   ü  
  70. 70   Unused  flexibility  

  71. 71   Illegal  variant  

  72. 72   72   Extensible  architectures   (plugins-­‐based)   ConfiguraCon

      files   System   Preferences   Configurators   Source  code   Build  systems   Comparison  of  Product  
  73.   Variability  AbstracCon  Model   (VAM)     CommunicaCve  

      AnalyCc     GeneraCve     73   not, and, or, implies
  74. Variability  Model   Feature  Model:  de  facto  standard   • 

    Research     –  2500+  citaWons  of  [Kang  et  al.,  1990]  on  Google  Scholar     –  Central  to  many  generaWve  approaches   •  at  requirements  or  code  level   –  Tools  &  Languages  (GUIDSL/FeatureIDE,  SPLOT,  FaMa,   etc.)   •  Industry     –  Tools  (Gears,  pure::variants),     •  Common  Variability  Language  (CVL)   74  
  75. None
  76. Feature  Models   76  

  77. Feature  Models  (Background)   77   Hierarchy:  rooted  tree  

      Variability:     •  mandatory,     •  opWonal,     •  Groups:  exclusive  or  inclusive  features   •  Cross-­‐tree  constraints   Optional Mandatory Xor-Group Or-Group
  78. 78   Hierarchy  +  Variability     =    

    set  of  valid  configuraCons   {CarEquipment,  Comfort,  DrivingAndSafety,  Healthing,  AirCondiWoning,  FrontFogLights}   configuraCon  =  set  of  features  selected   Optional Mandatory Xor-Group Or-Group
  79. 79   Hierarchy  +  Variability     =    

    set  of  valid  configuraCons   {CarEquipment,  Comfort,  DrivingAndSafety,  Healthing,  AirCondiWoning}   configuraCon  =  set  of  features  selected   Optional Mandatory Xor-Group Or-Group
  80. 80   Hierarchy  +  Variability     =    

    set  of  valid  configuraCons   Optional Mandatory Xor-Group Or-Group {CarEquipment,  Comfort,  DrivingAndSafety,  Healthing,  AirCondiWoning,   AutomaWcHeadLights}   configuraCon  =  set  of  features  selected   ü   ü   ü   ü   ü   ü  
  81. 81   Hierarchy  +  Variability     =    

    set  of  valid  configuraCons   Optional Mandatory Xor-Group Or-Group {AirCondiWoning,  FrontFogLights}   {AutomaWcHeadLights,  AirCondiWoning,  FrontFogLights}   {AutomaWcHeadLights,  FrontFogLights,  AirCondiWoningFrontAndRear}   {AirCondiWoningFrontAndRear}   {AirCondiWoning}   {AirCondiWoningFrontAndRear,  FrontFogLights}   {CarEquipment,  Comfort,   DrivingAndSafety,   Healthing}   X
  82. Feature  Models   82  

  83. None
  84.  (FeAture  Model  scrIpt  Language  for  manIpulaWon  and  AutomaWc  Reasoning)  

      not, and, or, implies φ TVL DIMACS hip://familiar-­‐project.github.com/   Mathieu  Acher,  Philippe  Collet,  Philippe  Lahire,  Robert  B.  France  «  A  Domain-­‐Specific  Language  for  Large-­‐ Scale  Management  of  Feature  Models  »  Science  of  Computer  Programming  (SCP),  2013  
  85. 85   Optional Mandatory Xor-Group Or-Group

  86. 86   Optional Mandatory Xor-Group Or-Group

  87. 87   Optional Mandatory Xor-Group Or-Group {AirCondiWoning,  FrontFogLights}   {AutomaWcHeadLights,

     AirCondiWoning,   FrontFogLights}   {AutomaWcHeadLights,  FrontFogLights,   AirCondiWoningFrontAndRear}   {AirCondiWoningFrontAndRear}   {AirCondiWoning}   {AirCondiWoningFrontAndRear,  FrontFogLights}   {CarEquipment,  Comfort,   DrivingAndSafety,   Healthing}   X
  88. 88  

  89. None
  90.  (FeAture  Model  scrIpt  Language  for  manIpulaWon  and  AutomaWc  Reasoning)  

      imporCng,  exporCng,  composing,  decomposing,  ediCng,  configuring,   reverse  engineering,  compuCng  "diffs",  refactoring,  tesCng,     and  reasoning  about  (mulCple)  variability  models   not, and, or, implies φ TVL DIMACS hip://familiar-­‐project.github.com/   Mathieu  Acher,  Philippe  Collet,  Philippe  Lahire,  Robert  B.  France  «  A  Domain-­‐Specific  Language  for  Large-­‐ Scale  Management  of  Feature  Models  »  Science  of  Computer  Programming  (SCP),  2013  
  91. #1  Automated  Analysis   91  

  92. #2  MulCple  Feature  Models   92  

  93. 93   93   MulC-­‐*  variability        

      *systems,  perspecCves,  or  stakeholders  
  94. •  #1  Automated  analysis     –  Aka  support  to

     beber  understand  and  play  with  your  feature   model  (TVL  model)   •  #2  Managing  mulCple  feature  models   –  Composing  /  Decomposing  /  Diff  and  Reasoning  about  their   relaWonships   –  Combining  these  operators   94   Two  Key  Requirements  
  95. language  and  environment              

      And-Group Optional Mandatory Xor-Group Or-Group constraints …….. DirectX V10 V10.1 v11 Outputs VIVO DVI HDMI S-Video Composite VGA GraphicCard And-Group Optional Mandatory Xor-Group Or-Group TV output constraints VGA excludes TV output HDMI implies v10.1 or v11 constraints …….. constraints …….. constraints …….. //  foo.fml   fm1  =  FM  (“foo1.tvl”)   fm2  =  FM  (“foo2.m”)   fm3  =  merge  intersecCon  {  fm1  fm2  }   c3  =  counCng  fm3   renameFeature  fm3.TV  as  “OutputTV”   fm5  =  aggregate  {  fm3  FM  (“foo4.xml”)  }   assert  (isValid  fm5)       fm6  =  slice  fm5  including  fm5.TV.*     export  fm6     True/False   8759   “OutputTV”,  “TV”     Interoperability   Language  faciliCes   Environment  
  96. 96   Interoperability   fm1  =  FM(“foo.tvl”)   fm2  =

     FM  (“foo.m”)     serialize  fm4  into  SPLOT   serialize  fm1  into  featureide   fm3  =  FM  (“foo.xmi”)   fm4  =  FM  (A  :  B  ….)        De/ComposiCon   merge                        diff                        intersecWon                        sunion       aggregate    map    unmap   extract                                                      slicing   EdiCng   renameFeature    removeFeature   accessors      copy                 Reasoning     counWng   configs   isValid   deads   cores   falseOpWonals   cleanup   configuraWon      select    deselect    asFM   compare   setOpWonal                          setMandatory   setAlternaWves      setOr      Language  FaciliCes   fm1.*   fm1.B   modular  mechanisms       restricted  set  of  types   iterator/condiWonal   asserWon   insert   features  
  97. Hello  World   97   helloworld.fml   Optional Mandatory Xor-Group

    Or-Group
  98. Typed  language     •  Domain-­‐specific  types   –  Feature

     Model,     –  ConfiguraWon,     –  Feature,     –  Constraint     •  Other  types  include     –  Set   –  String     –  Boolean,     –  Enum,     –  Integer  and  Real.     •  A  set  of  operaWons,  called  operators,  are  defined  for  a  given  type.     98   basics2.fml  
  99. Typed  language     99   basics2.fml  

  100. Typed  language     100   basics2.fml   Optional Mandatory

    Xor-Group Or-Group
  101. ImporCng/ExporCng  feature  models   101   FAMILIAR S2T2 TVL feature-model-synthesis

    (visual configurator) (language) (language) FaMa Internal  notaWon  or  by  “filename  extensions”     basics3.fml  
  102. Feature  Accessors  (1)   102   6Accessors.fml   Optional Mandatory

    Xor-Group Or-Group
  103. Other  constructs   103   6Accessors2.fml   Optional Mandatory Xor-Group

    Or-Group
  104. ConfiguraCon   104   conf.fml   Optional Mandatory Xor-Group Or-Group

  105. 105   φ FM A  ^   A  ó  B

     ^     C  =>  A  ^   D  =>  A     Optional Mandatory Xor-Group Or-Group
  106. OperaCons  for  Feature  Models  (1)   106   φ operatorsFM.fml

      Optional Mandatory Xor-Group Or-Group
  107. OperaCons  for  Feature  Models  (2)   107   φ operatorsFM2.fml

      Optional Mandatory Xor-Group Or-Group
  108. OperaCons  for  Feature  Models  (3)   108   operatorsFM3.fml  

    Optional Mandatory Xor-Group Or-Group
  109.                    

                                                                SoC  support  =  ComposiCon/DecomposiCon   for  managing   large,  complex  and  mulCple   feature  models   FORM  1998,  Tun  et  al.  2009  (SPLC),  Hartmann  2008  (SPLC),  Lee  et  al.  2010,  Czarnecki  2005,  Reiser  et  al.  2007  (RE  journal),  Hartmann   et  al.  2009  (SPLC),  Thuem  et  al.  2009  (ICSE),  Classen  et  al.  2009  (SPLC),  Mendonca  et  al.  2010  (SCP),  Dunghana  et  al.  2010,  Hubaux  et   al.  2011  (SoSyM),  Zaid  et  al.  2010  (ER),  She  et  al.,  2011  (ICSE),  etc.  
  110. Composing  Feature  Models  (1)   110   aggregateBasics.fml   Optional

    Mandatory Xor-Group Or-Group
  111. Composing  Feature  Models  (2)   111   aggregate1.fml   Previous

      version   Optional Mandatory Xor-Group Or-Group
  112. Composing  Feature  Models  (3)   112   mergeMI.fml   Mathieu

     Acher,  Philippe  Collet,  Philippe  Lahire,  Robert  B.  France  «  Comparing  Approaches  for   ImplemenWng  Feature  Model  ComposiWon  »  ECMFA’10  
  113. see  also  Thuem,  Kastner  and  Batory,  ICSE’09   Comparing  Feature

     Models   113   compare.fml   Optional Mandatory Xor-Group Or-Group
  114. None
  115. Merge  IntersecCon:  Available  Suppliers   115   ∩   ∩

      A  customer   has  some   requirements   Suppliers?   Products?  
  116. In  FAMILIAR   116   suppliersExample0.fml  

  117. Merge  Union:  Availability  Checking   117   Can  suppliers  provide

     all  products?   Yes!   “compare”         ∩   Optional Mandatory Xor-Group Or-Group
  118. In  FAMILIAR   118   suppliersExample.fml  

  119. Merging  operaCon:    implementaCon  issues   119   How  to

     synthesise  a  feature  model  that  represents   the  union  of  input  sets  of  configuraCons?   Optional Mandatory Xor-Group Or-Group T2 MRI Medical Image Header Anonymized T1 DICOM Header excludes DICOM Header implies Anonymized Anonymized v Header v ~DICOM v ~T1 v ~T2 Anonymized v Header v DICOM v ~T1 v ~T2
  120. 120   Merging  operaCon:  semanCc  issues  (2)   φ Union

      IntersecWon     Diff     How  to  synthesise  a  feature  model  that  represents   the  union  of  input  sets  of  configuraCons?  
  121. Merging  operaCon:  algorithm   121   φ 1 φ 2

    φ 3 φ 123 merged  proposiWonal  formula   T2 MRI Medical Image Header Anonymized T1 DICOM merged  hierarchy   +   Set  mandatory  features   Detect  Xor  and  Or-­‐groups   Compute  “implies/excludes”   constraints   How  to  synthesise  a  feature   model  that  represents  the   union  of  input  sets  of   configuraCons?   see  also  [Czarnecki  SPLC’07  or  SPLC’12]   Optional Mandatory Xor-Group Or-Group
  122. Merging  operaCon:  back  to  hierarchy   122   mergeNonPC.fml  

    >  configs  fm4   res12:  (SET)  {{C;A};{A;B};{A};{A;B;C}}   ?   Mathieu  Acher,  Benoit  Combemale,  Philippe  Collet,  Olivier  Barais,  Philippe  Lahire,  Robert  B.   France  «  Composing  your  ComposiWons  of  Variability  Models  »  MODELS’13   Optional Mandatory Xor-Group Or-Group
  123. see  also  [Acher  et  al.,  ECMFA’10  /  MODELS’13]   – Well-­‐defined

     semanWcs   – Guarantee  semanWcs  properWes  by  construcWon   – More  compact  feature  models  than  reference-­‐based   techniques  [Schobbens  et  al.,  2007],  [Hartmann  et  al.,  2007]   •  Easier  to  understand   •  Easier  to  analyze  (e.g.,  compare  with  another)   – Applicable  to  any  proposiWonal  feature  models     •  Full  support  of  proposiWonal  constraints     •  Different  hierarchies  [Van  Den  Broek  et  al.,  SPLC’2010/2012]   – SyntacWcal  strategies  fail  [Alves  et  al.,  2006],  [Segura  et  al.,  2007]   123   Related  Works  
  124. None
  125. 125   Problem:  mulCple  „car  models“    

  126. 126   Problem:  mulCple  „car  models“    

  127. 127   Problem:  mulCple  „car  models“    

  128. 128   Problem:  mulCple  „car  models“       #2

     –  boiom-­‐up:  elaborate  a  feature  model  for  each  model  line  and  merge  them   Two  modeling  approaches   #1  –  top-­‐down:  specify  constraints  (e.g.,  excludes)  of  all  model  lines  upfront    
  129. 129   #1  top-­‐down  

  130. 130   #1  boiom-­‐up   FM_1   FM_2   FM_3

      FM_r   merge  
  131. 131   #1  boiom-­‐up  (FAMILIAR)   FM_1   FM_2  

    FM_3   FM_r   merge   audiMerge.fml  
  132. None
  133. 133   Building  “views”  of  a  feature  model  

  134. •  Problem:  given  a  feature  model,  how  to   decompose

     it  into  smaller  feature  models?   •  SemanWcs?   – What’s  the  hierarchy   – What’s  the  set  of  configuraWons?   134   Building  “views”  of  a  feature  model  
  135. A  first  try   A3 => P1 P2 => A5

    R A2 A5 A6 A1 A3 A4 A fm0 P3 P2 P1 P P1 => P2 A2 A5 A6 A1 A3 A4 A fmExtraction1 A2 A5 A6 A1 A3 A4 A fmExtraction2 A3 => A5 A4 => A6 Problem:  You  can  select  A3  without  A5   Hierarchy  and  ConfiguraCon  maier!   135  
  136. Slicing  Operator   W constraints E implies D R implies

    E D excludes F S implies (F and not E) P R S fm1 A V T U B C D E F Optional Mandatory Xor-Group Or-Group T S E D constraints E implies D D implies E slicing  criterion  :  an  arbitrary  set  of  features,  relevant  for  a  feature  model  user   slice  :  a  new  feature  model,  represenWng  a  projected  set  of  configuraWons     136  
  137. Slicing  operator:  going  into  details   projected  set  of  configuraCons

      137   fm1  =  {     {A,B,C,D,E,P,R,T,U,W},     {A,B,C,F,P,S,T,U,W},     {A,B,C,D,E,P,R,T,W},     {A,B,C,F,P,S,T,V,W},     {A,B,C,F,P,S,T,U,V,W},     {A,B,C,F,P,S,T,W},     {A,B,C,D,E,P,R,T,V,W},     }   fm1  =  {     {A,B,C,D,E,P,R,T,U,W},     {A,B,C,F,P,S,T,U,W},     {A,B,C,D,E,P,R,T,W},     {A,B,C,F,P,S,T,V,W},     {A,B,C,F,P,S,T,U,V,W},     {A,B,C,F,P,S,T,W},     {A,B,C,D,E,P,R,T,V,W},     }   fm1p  =  {     {D,E,T},     {S,T},     {D,E,T},     {S,T},     {S,T},     {S,T},     {D,E,T}   }   fm1p  =  {     {D,E,T},     {S,T},     }   W constraints E implies D R implies E D excludes F S implies (F andnot E) P R S fm1 A V T U B C D E F Optional Mandatory Xor-Group Or-Group
  138. +   T S E D constraints E implies D

    D implies E φ s1 existenBal   quanBficaBon   of  features   not  included   in  the  slicing   criterion   138   fm1p  =  {     {D,E,T},     {S,T}   }   Slicing  operator:  going  into  details   synthesizing  the  corresponding  feature  model   S   E   D   T   W constraints E implies D R implies E D excludes F S implies (F andnot E) P R S fm1 A V T U B C D E F φ 1 Mathieu  Acher,  Philippe  Collet,  Philippe  Lahire,  Robert  B.  France  «  SeparaWon  of  Concerns  in   Feature  Modeling:  Support  and  ApplicaWons  »  AOSD’12   Optional Mandatory Xor-Group Or-Group
  139. T S E D constraints E implies D D implies

    E 139   Slicing  operator  with  FAMILIAR  (1)   W constraints E implies D R implies E D excludes F S implies (F andnot E) P R S fm1 A V T U B C D E F slicingOp2.fml   Optional Mandatory Xor-Group Or-Group
  140. 140   Slicing  with  FAMILIAR  (2)   slicingOp.fml  

  141. From  marke.ng,   customers,  product   management     From

     exis.ng  so@ware   assets    (technical  variability)   Metzger,  Heymans  et  al.  “DisambiguaBng  the  DocumentaBon  of  Variability  in  Sofware  Product   Lines:  A  SeparaBon  of  Concerns,  FormalizaBon  and  Automated  Analysis“  (RE’07)  
  142. V1 ⬄ f1 V2 ⬄ f2 V3 ⬄ f3 From

     marke.ng,   customers,  product   management     From  exis.ng   so@ware  assets     realizability   usefulness   Optional Mandatory Xor-Group Or-Group
  143. Realizability  checking   aggregate   {{V1,V3,V2,VP1},   {V1,VP1},   {V3,VP1},

        {VP1}}     merge  diff   (“unrealizable  products”   φ 1 slice  (“realizable  part”)   2 3 compare   4   Mathieu  Acher,  Philippe  Collet,  Philippe  Lahire,  Robert  B.  France  «  SeparaWon  of  Concerns  in   Feature  Modeling:  Support  and  ApplicaWons  »  AOSD’12     Optional Mandatory Xor-Group Or-Group
  144. With  FAMILIAR   144   realizibility.fml  

  145. None
  146. 146   RevisiCng  Merge:     Aggregate  +  Slice  

    Optional Mandatory Xor-Group Or-Group
  147. 147   RevisiCng  Aggregate,     Merge  and  Slice:  

        mergeWithAggregateMI.fml   Mathieu  Acher,  Benoit  Combemale,  Philippe  Collet,  Olivier  Barais,  Philippe  Lahire,  Robert  B.   France  «  Composing  your  ComposiWons  of  Variability  Models  »  MODELS’13   Optional Mandatory Xor-Group Or-Group
  148. 148   Mathieu  Acher,  Benoit  Combemale,  Philippe  Collet,  Olivier  Barais,

     Philippe  Lahire,  Robert  B.   France  «  Composing  your  ComposiWons  of  Variability  Models  »  MODELS’13  
  149. 149   φ FM            

    Feature  Model  Synthesis  Problem   [Czarnecki  et  al.,  SPLC’07]   [She  et  al.,  ICSE’11]   [Andersen  et  al.,  SPLC’12]   A  ^   A  ó  B  ^     C  =>  A  ^   D  =>  A    
  150. φ               « How

    to synthesise an accurate (w.r.t. the set of constraints/configurations) meaningful (maintainable by a user), and unique feature model? » hip://familiar-­‐project.github.com/  
  151. φ (SAT solvers or Binary Decision Diagrams) The  knowledge  can

     be:       inconsistent  (e.g.,  root  feature  specified  is  not  possible)   consistent  and  incomplete  (i.e.,  synthesis  algorithm  needs   addiWonal  informaWon)   consistent,  «  parCal  »  (e.g.,  not  all  the  hierarchy  is  specified)  and   actually  complete     Mathieu  Acher,  Patrick  Heymans,  Anthony  Cleve,  Jean-­‐Luc  Hainaut,  Benoit  Baudry  «  Support  for   Reverse  Engineering  and  Maintaining  Feature  Models  »  VaMoS’13  
  152. #1  Reverse  Engineering  Scenarios   •  [Haslinger  et  al.,  WCRE’11],

     [Acher  et  al.,  VaMoS’12]   φ V D Ad O T M K Ae C P R S C requires T Ae requires T S equals M V D Ad O T K Ae S P R M C requires T S equals M C 0..1
  153. #2  Refactoring   •  [Alves  et  al.,  GPCE’06],  [Thuem  et

     al.,  ICSE’09]   φ V D Ad O T M K Ae C P R S C requires T Ae requires T S equals M
  154. #3  Re-­‐Engineering  Feature  Models  of          

    repository   •  For  each  FM  we  execute  the  following  FAMILIAR  script…     •  …  And  we  «compare»  syntacWcally  fm1  and  fm2   •  semanWcal  comparison  is  not  needed:  we  know  that  they  are  refactoring  by   construcWon  (good  test  case  though  ;-­‐))   •  Results:   –  147  synthesised  FMs  (69  %)  were  exactly  the  same  as  input  FMs  ;     –  40  synthesised  FMs  (19%)  were  correcWons  of  input  FMs  ;     –  24  synthesised  FMs  (12%)  were  different  (knowledge  needed)   •  another  set  of  cross-­‐tree  constraints  was  synthesised.     •  feature  group  conflicts  in  six  cases   SpecificaCon  of  the  hierarchy  is  the  main  issue   φ
  155. φ             FAMILIAR   «

    Give me a formula and some knowledge, I will synthesise an accurate, meaningful, unique feature model » #1  Breathing  knowledge  into  feature  model  synthesis    formal  specificaWon  (consistency  and  completeness)    concrete  syntax  and  tooling  suport   #2  PracCcal  applicaCons    reverse  engineering,  refactoring/re-­‐engineering  of  feature  models       hip://familiar-­‐project.github.com/   Automated  support  is  highly   needed  (ongoing  work)  
  156. [MOTIVATION/PROBLEM]  Why  modeling  and  managing  Variability   does  and  will

     maber  (30’)   [SOLUTION  FOR  MANAGING  FEATURE  MODELS]  Managing  Variability   Models  with  FAMILIAR  (1h45’)       [SOLUTION  FOR  MODEL-­‐BASED  DERIVATION  OF  PRODUCT]  Model-­‐based   variability  engineering:  examples,  support  and  open  issues   (45’)   156   Plan  
  157. Variability  AbstracCon   Model  (VAM)   ConfiguraCon   (resoluCon  model)

      Domain  Artefacts   (e.g.,  models)   SoSware  Generator   (derivaCon  engine)   ü   ü   Variability   RealizaCon   Model   (VRM)  
  158. Configurations Derivation Process Models of the “system” Feature Model How

    to realize the variability
  159. Printer ´Blockª mainSupply:MainPower 1 Attributes Operations powerCtrl emgSupply:EmgPower 1 Attributes

    threshold:int Operations powerCtrl inputSection 1 highSpeedConnector 1 Attributes Operations MainPowerCtrl EmgPowerCtrl MainPower ´Blockª Values Operations powerCtrl EmgPower ´Blockª Values threshold:int powerCtrl VariaCon  Points  over  base  model   :ObjectExistence   :SlotValueAssignment   CVL variation points SYSML (base model) elements :ObjectExistence   :ObjectExistence   §  Variability  in  this  example:     Part  EmergencySupply  is   opWonal     Part  HighSpeedConnector  is   opWonal     Port  EmgPowerCtrl  on  block   Printer  is  opWonal     Value  of  abribute  threshold  in   block  EmergencyPower  is   variable   Adapted  from  the  CVL  tutorial  at  SPLC’12  by  Oystein  Haugen,  Andrezj  Wasowski,  Krzysztof   Czarnecki    
  160. Printer ´Blockª mainSupply:MainPower 1 Attributes Operations powerCtrl emgSupply:EmgPower 1 Attributes

    threshold:int Operations powerCtrl inputSection 1 highSpeedConnector 1 Attributes Operations MainPowerCtrl EmgPowerCtrl MainPower ´Blockª Values Operations powerCtrl EmgPower ´Blockª Values threshold:int powerCtrl (Aiributed)   Feature  Model     Printer   EmergencyPower   threshold:Int   Variation points HighSpeed  &    threshold>100      EmergencyPower   HighSpeed   :ObjectExistence   :SlotValueAssignment   :ObjectExistence   :ObjectExistence   Based  Model     Adapted  from  the  CVL  tutorial  at  SPLC’12  by  Oystein  Haugen,  Andrezj  Wasowski,  Krzysztof   Czarnecki    
  161. Variability  RealizaCon  Layer   VariaCon  points  in  CVL   • 

    VariaWon  Points  refer  to  Base  objects   •  VariaWon  Points  define  the  base  model   modificaWons  precisely   •  There  are  different  kinds  of  VariaWon  Points   – Existence  (object  or  link)   – Value  assignment   – SubsWtuWon   – Opaque  variaWon  point   – Configurable  Unit   Adapted  from  the  CVL  tutorial  at  SPLC’12  by  Oystein  Haugen,  Andrezj  Wasowski,  Krzysztof   Czarnecki    
  162. Derivation of Traffic Lights Models Joao  Bosco  Ferreira   Filho

     (PhD  student)  
  163. Traffic Lights

  164. . Traffic Lights' behaviour can be modelled using Finite State

    Machines Traffic Lights FSM
  165. Traffic Lights FSM OBJECTIVE: Produce the finite state machine associated

    with each traffic light configuration . Simplification: - Green Light - Red Light - Yellow Light
  166. Base model . Define the DSL: 1. Create an Ecore

    modelling project 2. Define the metamodel 3. Generate the model, the edit and the editor code 4. Export them as plugins
  167. DSL metamodel

  168. Base model . Create the base model: 1. Create a

    Modelling project 2. Create a new model in the DSL
  169. Base model . Create the base model: 1. Create a

    Modelling project 2. Create a new model in the DSL
  170. CVL model . Create the CVL model: 1. Create a

    new CVL model in the modelling project. Select VPackage as the Model object 2. Right click on the project, select Viewpoints selection. Check the three of them 3. Define the VAM, Resolution model and VRM
  171. VAM

  172. Resolution model

  173. Resolution model

  174. VRM

  175. VRM

  176. Product derivation

  177. Derived FSM

  178. Visualisation of products in a web configurator Marianela Ciolfi Felice

    (MSc student)  
  179. None
  180. Marks  &  Spencer  web  configurator  

  181. §  High visual quality §  Coherence and stability §  Interactiveness

    §  Performance §  Automatic and comprehensive update method Marianela Ciolfi Felice and Joao Bosco Ferreira Filho and Mathieu Acher and Arnaud Blouin and Olivier Barais « Interactive Visualisation of Products in Online Configurators: A Case Study for Variability Modelling Technologies » MAPLESCALE’13 (to appear)
  182. Anticipating all possible combinations §  10 configuration options §  10

    possible values for each of them 10.000.000.000 combinations! Composing the visualisation
  183. PRODUCT CONFIGURATION VISUAL REPRESENTATION Models   HTML   jpg,  png,

     ...,   files   SVG  files   Javascript   3D   models   Feature   models  
  184. Configurations Derivation Process Visual representation of a product Feature Model

    How to realize the variability Visual elements
  185. . Simplification: - Fabric - Collar - Pocket (optional) -

    Handkerchief (optional)     Shirts web configurator
  186. Shirts web configurator OBJECTIVE: Produce the visual representation associated with

    each shirt configuration . Simplification: - Fabric - Collar - Pocket (optional) - Handkerchief (optional)    
  187. Feature model - Implicit boolean attribute existence - No constraints

  188. Base model . Define the DSL: 1. Create an Ecore

    modelling project 2. Define the metamodel 3. Generate the model, the edit and the editor code 4. Export them as a plugin
  189. DSL metamodel

  190. Base model . Create the base model: 1. Create a

    Modelling project 2. Create a new model in the DSL
  191. Base model . Create the base model: 1. Create a

    Modelling project 2. Create a new model in the DSL
  192. CVL model . Create the CVL model: 1. Create a

    new CVL model in the modelling project. Select VPackage as the Model object 2. Right click on the project. Select Viewpoints selection. Check the three of them 3. Define the VAM, Resolution model and VRM
  193. VAM

  194. VAM

  195. VAM Suggestion: Set the choices' default resolution to: - True

    for mandatory features - False for optional features
  196. Resolution model

  197. Resolution model

  198. VRM

  199. VRM

  200. Product derivation

  201. Product derivation

  202. Summary:  Variability  Model  Management   202   202   202

     
  203. [MOTIVATION/PROBLEM]  Why  modeling  and  managing  Variability   does  and  will

     maber  (30’)   [SOLUTION  FOR  MANAGING  FEATURE  MODELS]  Managing  Variability   Models  with  FAMILIAR  (1h45’)       [SOLUTION  FOR  MODEL-­‐BASED  DERIVATION  OF  PRODUCT]  Model-­‐based   variability  engineering:  examples,  support  and  open  issues   (45’)   203   Key  Insights  
  204. (ongoing)   Comprehensive  model-­‐based   product  line  support    

    Reverse  engineering   Automated  Analysis   Languages,  API/DSLs   EvaluaCon  (European  projects,  long-­‐term  collaboraCon  with  Thales,  open  source   systems)     204   204  
  205.                    

    ?   205