UML 2 0 Roadmap What Should Users Expect

  • Slides: 23
Download presentation
UML 2. 0 Roadmap: What Should Users Expect? Cris Kobryn Cris. Kobryn@telelogic. com Chief

UML 2. 0 Roadmap: What Should Users Expect? Cris Kobryn Cris. Kobryn@telelogic. com Chief Technologist, Telelogic UML 2. 0 Roadmap

Overview n n n What is UML? Why is UML important? Why UML 2.

Overview n n n What is UML? Why is UML important? Why UML 2. 0? UML 2. 0 status Featured proposal: U 2 Partners’ submission Issues UML 2. 0 Roadmap 2

What is UML? n Unified Modeling Language is a graphical language for n n

What is UML? n Unified Modeling Language is a graphical language for n n n the artifacts of software systems Derived from synthesis of Booch, OMT and Objectory modeling languages n n specifying visualizing constructing documenting roots can be traced to data modeling and SDL Adopted by OMG in November 1997 as UML 1. 1 n n most recent minor revision is UML 1. 4, adopted in May 2001 next planned major revision is UML 2. 0, scheduled to be completed during 2002 UML 2. 0 Roadmap 3

Why is UML important? n As software architectures grow in size and complexity so

Why is UML important? n As software architectures grow in size and complexity so does the need for software models n n UML is the software industry’s dominant modeling language n n compare building blueprints and visual CAD currently a de facto standard adopted by the Object Management Group, the world’s largest software consortium may soon be a de jure international standard (OMG is submitting to ISO as Publicly Available Specification) difficult to find a software project with > 10 developers who do not use UML in some way to specify their architecture UML has enormous growth potential n n lingua franca across the software lifecycle (requirements through testing) and across platforms and domains executable UML can automate software development reducing the chasm between analysis & design and implementation UML 2. 0 Roadmap 4

Why UML 2. 0? n Although UML 1. x has enjoyed widespread acceptance, it

Why UML 2. 0? n Although UML 1. x has enjoyed widespread acceptance, it shortcomings include: n n n n n excessive size gratuitous complexity uneven precision limited customizability inadequate support for components non-standard implementations lack of support for diagram interchange … A major revision is required to address these issues UML 2. 0 Roadmap 5

UML 2. 0 Status n n n Requests for Proposals (RFPs) Proposals Schedule UML

UML 2. 0 Status n n n Requests for Proposals (RFPs) Proposals Schedule UML 2. 0 Roadmap 6

UML 2. 0 RFPs Diagram Interchange Superstructure Defines user-level constructs to model structure and

UML 2. 0 RFPs Diagram Interchange Superstructure Defines user-level constructs to model structure and behavior of systems (e. g. , class diagrams, etc. ) Object Constraint Language Infrastructure Defines basic constructs to specify UML and to customize it for specific domains (e. g. , real-time) UML 2. 0 Roadmap 7

UML 2. 0 Submissions n UML 2. 0 Infrastructure n n UML 2. 0

UML 2. 0 Submissions n UML 2. 0 Infrastructure n n UML 2. 0 Superstructure n n 37 LOIs; 5 initial submissions by 28 companies UML 2. 0 OCL n n 36 Letters of Intent (LOIs); 5 initial submissions by 28 companies 30 LOIs; 4 initial submissions by 10 companies UML 2. 0 Diagram Interchange n 6 LOIs; 3 initial submissions by 6 companies UML 2. 0 Roadmap 8

UML 2. 0 Schedule 26 Apr 02 UML 1. 4. 1 final report 3

UML 2. 0 Schedule 26 Apr 02 UML 1. 4. 1 final report 3 Jun 02 UML 2 Infrastructure & OCL revised submissions Feb Mar Apr May Jun Jul 19 Aug 02 UML 2 Superstructure & Diagram Interchange revised submissions 25 Nov 02 UML 1. x Action Semantics final report Aug Jan 2002 Sep Oct Nov Dec 2002 UML 2. 0 Roadmap 9

Featured Proposal: U 2 Partners n Insufficient time to review all proposals n n

Featured Proposal: U 2 Partners n Insufficient time to review all proposals n n please download all submissions and judge for yourself U 2 Partners proposal selected because n n n work of the largest submission team (28 submitters and supporters) balanced approach to integrating language Infrastructure with Superstructure architecturally aligned with other industry standards, such as ITU-T languages UML 2. 0 Roadmap 10

U 2 Partners’ Proposal n A consortium of UML vendors and users dedicated to

U 2 Partners’ Proposal n A consortium of UML vendors and users dedicated to making UML easier to apply, implement and customize n Submitters n n Alcatel, CA, ENEA, Ericsson, HP, IBM, I-Logix, IONA, Kabira, Motorola, Oracle, Rational, SOFTEAM, Telelogic, Unisys, Web. Gain Supporters n Ceira, Commissariat à L'Energie Atomique, Embarcadero, Gentleware, Intellicorp, Jaczone, Kennedy Carter, Lockheed Martin, MSC. Software, Sims Associates, Syntropy, University of Kaiserslautern UML 2. 0 Roadmap 11

Goals n Restructure and refine the language to make it easier to apply, implement

Goals n Restructure and refine the language to make it easier to apply, implement and customize n n n increase precision practice reuse Infrastructure goals n n Define a metalanguage kernel that can define (bootstrap) UML and also be reused to define other OMG Model-Driven Architecture metamodels (e. g. , Meta Object Facility, Common Warehouse Model) Provide more powerful mechanisms to customize UML n allow users to define language dialects for platforms (e. g. , J 2 EE, . NET) and domains (e. g. , telecom, finance, mil/aero). UML 2. 0 Roadmap 12

Goals: Superstructure n Improve support for component-based development n n n Activities will be

Goals: Superstructure n Improve support for component-based development n n n Activities will be revised to support more flexible parallelism and furnish more I/O options. Increase scalability and precision of other behavioral constructs. n n State machines will be updated to make them more scalable and generalizable, and able to support a transition-centric view. Improve support for business process modeling. n n support hierarchical composition of parts with interfaces (compare SDL blocks and processes). Enhance support for real-time development. n n specify both platform-independent components (e. g. , business components) and platform-specific components (e. g. , EJB, COM+) Refine architectural specification capabilities. n n (cont’d) Sequence diagrams will adapt concepts from MSCs that can be combined and integrated with other behavior. Deepen precision so that it better supports executable models. Review all UML 1. x constructs and diagrams. n Refine, retire or deprecate as appropriate UML 2. 0 Roadmap 13

Cross-Pollination of Standards SDL MSC UML Other modeling languages UML 2. 0 Roadmap 14

Cross-Pollination of Standards SDL MSC UML Other modeling languages UML 2. 0 Roadmap 14

Coming Attractions n Examples of proposed notation for new UML 2. 0 constructs n

Coming Attractions n Examples of proposed notation for new UML 2. 0 constructs n work in progress UML 2. 0 Roadmap 15

Structured Classes Vending. Machine part : Coin. Handler Insert. Coin Dispense. Coin Ctrl connector

Structured Classes Vending. Machine part : Coin. Handler Insert. Coin Dispense. Coin Ctrl connector port Hdlr : Controller Display port UML 2. 0 Roadmap 16

Components provided interface required interface component UML 2. 0 Roadmap 17

Components provided interface required interface component UML 2. 0 Roadmap 17

Sequence Diagrams sd Overview ref Validate. Coin ref Choose. Product [Select] ref opt [else]

Sequence Diagrams sd Overview ref Validate. Coin ref Choose. Product [Select] ref opt [else] Dispense. Product ref Give. Change UML 2. 0 Roadmap 18

Sequence Diagrams (cont’d) sd Trace sd Validate. Coin Vending. Machine ref Validate User :

Sequence Diagrams (cont’d) sd Trace sd Validate. Coin Vending. Machine ref Validate User : User Show(” 1. 00”) Coin(c) alt : Vending. Machine Coin(0. 25) Show(” 0. 75”) Show(s) Coin(0. 10) else Show(” 0. 65”) Reject. Coin() sd Validate Coin. Handler Controller Coin(c) UML 2. 0 Roadmap 19

Issues n Second language syndrome n n n Paradigm evolution n n compare building

Issues n Second language syndrome n n n Paradigm evolution n n compare building blueprints and MVC architectures requires refinement and traceability across views Action semantics integration n n compare natural language jargon, dialects and families will SDL and/or MSC profiles still be required? Multiple views n n reduce impedance between object and component paradigms Language reuse: libraries, profiles, and metamodels n n “Will UML 2. 0 Be Agile or Awkward, ” CACM, vol. 45, no. 1, Jan. 02. net weight loss = features added – features removed – 0. 5 * features deprecated full integration vs. optional compliance point Architectural alignment n relationship to OMG MDA, ITU-T languages, etc. UML 2. 0 Roadmap 20

Wrap Up n UML is used extensively inside and outside OMG n n n

Wrap Up n UML is used extensively inside and outside OMG n n n UML’s evolution is following a well-planned roadmap: n n lingua franca for OMG’s MDA initiative de facto standard de jure standard UML 1. 4 is a mature minor revision UML 1. 4. 1 will be a maintenance revision UML 2. 0 will be next planned major revision U 2 Partners’ submission for UML 2. 0 proposes many improvements that will benefit users n n more concise and precise improved support for component development, architectural specifications, business process modeling, and executable models UML 2. 0 Roadmap 21

Resources: Web n U 2 Partners n n OMG UML Resources n n www.

Resources: Web n U 2 Partners n n OMG UML Resources n n www. uml. org UML Forum n n n www. u 2 -partners. org www. uml-forum. com Contains links to the UML Revision Task Force and UML 2. 0 Working Group webs as well as other UML resources. UML Models and Methods column n www. telelogic. com/publications/uml_models/ UML 2. 0 Roadmap 22

Resources: Other n Mailing Lists n n uml 2 -wg@omg. org uml-rtf@omg. org u

Resources: Other n Mailing Lists n n uml 2 -wg@omg. org uml-rtf@omg. org u 2 p-info@yahoogroups. org Conferences & Workshops UML Forum/Tokyo 2002, Tokyo, March 26 -27, 2002. n UML Track at SD West 2002, San Jose, CA, Apr. 22 -26, 2002. n UML 2002, Dresden, Germany, Sep. 30 -Oct. 4, 2002. n OMG UML Workshop, San Francisco, Dec. 3 -6, 2002. mailto: cris. kobryn@telelogic. com? subject=question n n UML 2. 0 Roadmap 23