![OCL(s)
• Collection types in OCL
• Set [1,2,3,4]
• Bag [‘a’,’f’,’r’,’a’]
• Sequence [‘white’,’green’,’red’]
• Iterators in OCL
• Collection -> size()
• Collection -> sum()
• Collection -> collect(expression)
• Collection -> select(expression)](https://image.slidesharecdn.com/projectbackuprepositoryandavoidingrequirementscreep-200923172709/85/Project-backup-repository-and-avoiding-requirements-creep-16-320.jpg)
Project backup repository and avoiding requirements creep
- 1. PROJECT BACKUP REPOSITORY AVOIDING REQUIRMENTS CREEP By Aswin Vijayakumar
- 2. PROJECT BACKUP REPOSITORY REQUIRMENTS CREEP
- 3. REQUIRMENTS CREEPPROJECT BLASTOFF Requirements creep is seen as one of the major problems in enterprises when developing new software.
- 4. PROJECT BLASTOFF REQUIRMENTS CREEP
- 5. Project Backup Repository • Essential for the long run of the project and post project execution stages such as handover and audits. • Includes a Quality Gateway for the entry, process, integration and finalisation of requirements. • Backs-up the refined requirements. What are refined requirements ?
- 6. Agile Requirements Refinery Process by which requirements cover its pool and get refined through: I. Vision II. Theme III. Concept IV. Requirements Definition It has the same structure as the stakeholder map, with the Vision at its centre and the requirements definitions at the outer circle.
- 7. What can you do with ARR? You can manage Creep:- 1) Creating GO / NO-GO decisions within the ARR 2) Creating extension use cases 3) Trace a requirement from its child to parent and vice versa You can manage the Product and Sprint Backlog:- 1) Stages in ARR secure its content through Backlogs 2) The repetitive process of ARR manages to populate the ARR map
- 8. GO – NO/GO
- 9. GO – NO/GO This is a nogo product as the specifications do not match
- 10. G/O – NOGO
- 11. G/O – NOGO
- 12. Quality Gateway Company X produces A amount of Goods in a day. Upon the agreed amount of goods produced in a day, 2% of the products are rejected. Rejected ones are thrown away for recycling. The ones which pass through the Gateway are either Bought or Sold. There are several factors that decide what goes through that gateway. For product based implementations, it is quality assurance. For project based implementations, we write a fit criterion for every requirement. Gateway can be written for: (1) Refinery, (2) Fit Criterion, (3) Traceability, (4) Deployment
- 13. USE CASE EXTENSIONS (a) (b) OCL(s) Collection -> isBought AND Collection -> isGateway Collection -> isRecycled OR Collection -> isNotGateway
- 14. Use cases and OCL(s) • Use cases have extension use cases. • They use fixtures of data to conduct tests on Pre-conditions and Post- conditions. • Each pre-condition and post-condition must be satisfied with OCL(s) • OCL(s) can be written in any programming language which takes in input data and produces a Boolean state for every line of requirements.
- 15. OCL(s) • Pre-coded constraints in the form of logic and instructions to validate a use-case • Basic OCL types • Boolean • Integer • Real • String • Operations of OCL types • And,or,+,-,/,xor,abs(),floor(),toUpper(),concat()
- 16. OCL(s) • Collection types in OCL • Set [1,2,3,4] • Bag [‘a’,’f’,’r’,’a’] • Sequence [‘white’,’green’,’red’] • Iterators in OCL • Collection -> size() • Collection -> sum() • Collection -> collect(expression) • Collection -> select(expression)
- 17. OCL(s) • Operators in OCL • Object.oclIsTypeOf(t: oclType) • Object.oclIsNew() • Object.oclIsKindOf(t: oclType)
- 18. Validation for Quality Gateway • Validation is an important step for ongoing Sprint Backlog or preceding or succeeding Product Backlog. • In the cases where Quality needs to be maintained for requirements coming to the Product / Sprint Backlog and for those going to the Project Backup Repository, the Product Manager enters logic to sustain the Quality Gateway.
- 19. Validation for Quality Gateway • Quality Gateway itself is one subset of the entire Requirements Pool. The client declares how much refinery should happen in 1 Sprint. The criteria are:- 1) How much Refinery 2) All fit criteria are passed 3) How much Traceability
- 20. Using DAMA for Quality Gateway DOCUMENT ROUTING Create a Master Record set of Requirements, Filter them using Quality Gateway criteria, Send them to create the Golden Record set of Requirements, DOCUMENT DATASET Use FDD to resolve requirements, The golden set is obtain on the basis of what to deploy
- 21. Requirements Prototyping To convey about the intended product, it is best to share prototypes. Prototyping comes in a variety of methods, the two main classes involved are:- 1) Low-fidelity 2) High-fidelity
- 22. Low-fidelity vs High-fidelity Online Seminars, Workshops, Face to Face interviews are some of the effective methods to perform requirements gathering process. Prototypes capture essential requirements and convey the intended product in order to obtain the feedback.
- 23. Low-fidelity vs High-fidelity LOW-FIDELITY HIGH-FIDELITY Paper User Interface Not functional Functional Screens Convertible
- 24. Metrics in Project Backup Repository SATISFACTION DISSATISFACTION Floating Point Rating Floating Point Rating Pros and Cons Pros and Cons Refinery Coefficient Refinery Coefficient Deployment Status Gateway Status • Story Points used to evaluate Efficiency using Data Envelopment Analysis (DEA) • Decision Management Units (DMUs) and Gini Impurity to evaluate Relevance • SHAP (Shapely Additive Explanations) to evaluate Importance
- 25. Efficiency, Relevance and Importance DECISION MANAGEMENT UNITS IMPORTANCE