Differences between revisions 32 and 33
Revision 32 as of 2006-06-20 21:23:44
Size: 6798
Editor: t-indiv10-132
Comment:
Revision 33 as of 2006-06-20 21:30:57
Size: 9554
Editor: t-indiv10-132
Comment:
Deletions are marked like this. Additions are marked like this.
Line 107: Line 107:
[[Anchor(feedback)]]
Line 108: Line 109:
[[Anchor(team)]] You have learnt the knowledge and skills that help to express the structure and behavior of the software components in their design in a way that is understood by involved parties. You have developed an understanding and appreciation of what it means to master complexity and to communicate it with engineers.

We appreciate that all the teams did their best to actively follow the lectures and practice with the exercises. The method of acting out the UML diagrams instead of implementing them using programming worked remarkably well, thanks to the efforts that all of you have put in the module.

Note that in one week of time, it is impossible to grasp every detail of UML and to make the diagrams perfect. Hence we are not going to give feedback on your diagrams here, except the comments that were given during your presentations. Again we would emphasize that there is no diagram that is absolutely right or wrong. A UML diagram which helps to express what is in your mind is a good diagram. More accurate is better. More consistent is better. Whether it is a good design is yet another issue - but as all you have found, a formal specification helps to clarify the concepts and to find the design flaws.

As we have already pointed out, this model helps you to give you some taste of UML as a formal specification language and its use in a design process. The focus was not on the diagrams themselves. But we did not mean that you should not continue to explore object-oriented design and formal specification methods. As we all found out, combining formal specifications and acting out in design cycles (even small cycles in spiral design or agile design) can be a good method in a design process. To put this method into your collection of design gadgets, you still need to learn more and practice more. Come back to us if you need any further help.

Your reflections on this module are published here. Read the comments you have got from the users of your diagrams – the team that acted out your diagrams. Read carefully what the others thought about your specifications:

 * Team 1 (Stefania Nicolosi, Esra Günlü, Jeroen Berk, Ruud-Peter Lemmens): attachment:team1.pdf
 * Team 2 (Fabian Koopman, Sander Kruitwagen, Marcel Verbunt, Thomas Visser , Carl Megens, Michel Peeters): attachement:team2.pdf
 * Team 3 (We did not have team 3)
 * Team 4 (Michel van der Hoek, Leoni Hurkx, Victor Versteeg, Benjamin Voss, Ralph Zoontjens): attachment:team4.pdf
 * Team 5 (Sibrecht Bouwstra, Sander Kouwenberg, Bart-Jan van Putten, Jos Verbeek, Alice Verdonk): attachment:team5.pdf
 * Team 6 (Pei-Yin Chao, Bas Groenendaal, John Helmes, Philip Mendels, Rik Wesselink, Mehmet Yalvac): attachment:team6.pdf

Formal Software Specification (UML Modelling)

May, 2006

TableOfContents

1. Module Description

A product with sensors, actuators and network connections can offer an interesting, useful, or playful behaviour to its users and to the other products, systems and services to which it is connected. The ID Master takes responsibility for the creation of this behaviour. If the product isn’t stand-alone, neither is the designer. Whenever product behaviour is realised through computer software and protocols, the designer takes advantage of being an excellent communicator in these matters. In present-day software engineering, the Universal Modelling Language UML has become widely accepted. It contains “activity diagrams”, “use case diagrams”, “class diagrams”, “state charts” and “message sequence charts”. The knowledge and skills that students get by participating in this module will help them to express the structure and behaviour of the software components in their design in a way that is understood by third parties. Starting from elementary programming skills, which are a prerequisite, the student will develop an understanding and appreciation of what it means to master complexity. The scope is widened from small programs to real complex software systems. Although developing and maintaining such systems usually involves computer scientists as well, the ID Master will be well equipped to use UML and thus specify system structure and desired behaviour.

Jun Hu, Yuechen Qian, and Loe Feijs will work together to provide background information, explain UML and create a learning experience in which reading UML, writing UML, and creating software come together.

Philip Ross is invited to give a lecture on the mothodolgy of acting out and his experiences in applying this method in designing products.

2. Schedule

  • /!\ All morning sessions start at 9:30AM.

  • 29 May, morning
  • Think Different: Object orientation and Use cases by Jun, (beamer needed)

    • Object Orientation:
      • Presentation: attachment:OO.pdf
      • Handouts: attachment:OOHandouts.pdf
      • Reading: [http://www.objectcentral.com/oobook/oobook.htm Essense of Object Oriented Programming] Several chapters of this book by Wampler are available online. Chapter 2 has a very good overview of object-orientation.

    • UML and Use Case diagrams
      • Presentation: attachment:UML.pdf
      • Handouts: attachment:UMLHandouts.pdf
      • Reading: attachment:UseCaseDiagram.pdf

  • 29 May, afternoon
  • Introduction to the Project: Zoo of Tamagotchi animals by Jun, (beamer needed)

  • Use cases of the Zoo of Tamagotchi animals.

  • 30 May, morning
  • Static view in UML by Yuechen, (beamer needed)

    • attachment:UMLStaticView.ppt
  • 30 May, afternoon
  • OO Modeling of the Zoo of Tamagotchi animals, coached by Jun

  • 31 May, morning
  • Behaviors in UML by Yuechen, (beamer needed)

  • 31 May, afternoon
  • FMS of the Zoo of Tamagotchi animals, coached by Jun

  • 1 June, morning
  • Finalizing the specification, coached by Jun
  • 1 June, afternoon
  • 12:00pm: Software engineering by Loe, (beamer needed)

  • 1:00pm: Acting out by Philip Ross, (beamer needed)

  • 2:00pm: Specification Handover
  • Read the specification from another team
  • Implementing the specification by acting, coached by Jun

  • 2 June, morning
  • Implementing the specification by acting, coached by Jun

  • 2 June, afternoon: The show
  • Format:
    • Every team performs the specification received from another team

    • The specification team comments on the performance.

    • Open discussions
  • Time:
    • 12:00pm-14:30pm: Performances and discussions
    • Every team gets 20 minutes at least, 30 minutes maximum.
    • 14:30pm-15:00pm: wrap up.

3. Project

3.1. Format

Students are divided into teams. Every team gets the same task: Designing a system called Zoo of Tamagochi Animals:

  • The first half of the project would be the teams making and specifying the OO design of the system;
  • At the end of the first half of the project, the teams swap their specifications for implementation;
  • The second half of the project would be implementing the specifications received.

Don't worry. No Java. No C++. No programming at all.

We are not going to do real software implementations. We will implement the system by acting the specification out to show how the system should work according to the specification. Students can play the objects, showing their behaviors and the communication in between. Stage tools can also be used to represent objects, interfaces, and events etc. We will need some imagination and creativity.

3.2. Zoo of Tamagotchi animals

A preliminary description of the requirement:

  • All animals live in a zoo.
  • An animal has got a life after it is born. While the life goes on, the animal moves and sleeps (if it is still alive), eats (if it moves and finds food), grows (if it eats) until one day, it dies because of hunger, illness or age.
  • Every animal has got a body. Different animals look different because of their different bodies. Every animal has got two eyes and one mouth on its body. When eyes are pinched, animals scream and can be hurt.
  • There are male and female animals. When grownup males and females meet, they may fall into love and the love may result in baby animals.
  • Some animals are pets. Pets have names and they wear their name plates on their bodies. People (the users) take care of their pets and feed them with food.
  • People may play with their pets to keep their pets fit.
  • Some baby animals will be selected by people and they become pets. The rest are left free in the zoo and they have to strive for food and try to survive by themselves.
  • When a zoo is created, it is empty, until people get some animals from somewhere else (from shops, for example).
  • People may exchange their pet animals.
  • Dogs and cats will be our favorite animals for the time being. Dogs bark and cats miaow. Cats are scared of dogs and can be bitten by dogs. When big dogs bark, cats scream and start running away. When big dogs fall into sleep, cats start getting together and partying.
  • The zoo is open for other animals, including unknown ones. attachment:barkdifferent.jpg

4. Modelling tools

Anchor(feedback)

5. Feedback to and from the teams

You have learnt the knowledge and skills that help to express the structure and behavior of the software components in their design in a way that is understood by involved parties. You have developed an understanding and appreciation of what it means to master complexity and to communicate it with engineers.

We appreciate that all the teams did their best to actively follow the lectures and practice with the exercises. The method of acting out the UML diagrams instead of implementing them using programming worked remarkably well, thanks to the efforts that all of you have put in the module.

Note that in one week of time, it is impossible to grasp every detail of UML and to make the diagrams perfect. Hence we are not going to give feedback on your diagrams here, except the comments that were given during your presentations. Again we would emphasize that there is no diagram that is absolutely right or wrong. A UML diagram which helps to express what is in your mind is a good diagram. More accurate is better. More consistent is better. Whether it is a good design is yet another issue - but as all you have found, a formal specification helps to clarify the concepts and to find the design flaws.

As we have already pointed out, this model helps you to give you some taste of UML as a formal specification language and its use in a design process. The focus was not on the diagrams themselves. But we did not mean that you should not continue to explore object-oriented design and formal specification methods. As we all found out, combining formal specifications and acting out in design cycles (even small cycles in spiral design or agile design) can be a good method in a design process. To put this method into your collection of design gadgets, you still need to learn more and practice more. Come back to us if you need any further help.

Your reflections on this module are published here. Read the comments you have got from the users of your diagrams – the team that acted out your diagrams. Read carefully what the others thought about your specifications:

  • Team 1 (Stefania Nicolosi, Esra Günlü, Jeroen Berk, Ruud-Peter Lemmens): attachment:team1.pdf
  • Team 2 (Fabian Koopman, Sander Kruitwagen, Marcel Verbunt, Thomas Visser , Carl Megens, Michel Peeters): attachement:team2.pdf
  • Team 3 (We did not have team 3)
  • Team 4 (Michel van der Hoek, Leoni Hurkx, Victor Versteeg, Benjamin Voss, Ralph Zoontjens): attachment:team4.pdf
  • Team 5 (Sibrecht Bouwstra, Sander Kouwenberg, Bart-Jan van Putten, Jos Verbeek, Alice Verdonk): attachment:team5.pdf
  • Team 6 (Pei-Yin Chao, Bas Groenendaal, John Helmes, Philip Mendels, Rik Wesselink, Mehmet Yalvac): attachment:team6.pdf

JunHu: UmlMasterModule/0605 (last edited 2012-05-25 12:44:42 by JunHu)