Sofia ly 2003 2 Ju r e m

Sofia ly 2003 2 Ju r e m m Su ol o h c S IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert une 30 J e. XPERT Best practices Boyan Angelov Penko Ivanov Nemetschek Rila Solutions e. XPERT Best practices Bulgari a

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o Sch une 30 J e. XPERT in Nemetschek Boyan Angelov e. XPERT Best practices Sofia ly 2003 2 Ju Bulgari a

r e m m u S l o o h c S Nemetschek OOD IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a The Pilot Project - Aims Apply e. XPERT in real life project Gather experience Collect metrics Evaluate the approach e. XPERT Best practices

r e m m u S l o o h c S Nemetschek OOD IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a The Pilot Project - Scope • Facility Management system • New technologies – Microsoft. NET • Real life customer • Small team e. XPERT Best practices

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o Sch Sofia ly 2003 2 Ju une 30 J Bulgari a e. XPERT Application – Phases Productive Phase Big Bang Evaluation Phase Project Lifetime e. XPERT Best practices

Sofia ly 2003 2 Ju r e m m u S l o o ch S e. XPERT Application – Phase 1/3 une 30 J IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Big Bang Initial Phase • Execute all practices at once • Apply strict control over practices application • Evaluate practices • Modify bottleneck practices e. XPERT Best practices Bulgari a

Sofia ly 2003 2 Ju r e m m u S l o o ch S e. XPERT Application – Phase 2/3 une 30 J IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Productive Phase • Apply the approach • Identify key practices • Refine practices application • Produce software e. XPERT Best practices Bulgari a

Sofia ly 2003 2 Ju r e m m u S l o o ch S e. XPERT Application – Phase 3/3 une 30 J IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Evaluation Phase • Collect quantitative measures • Evaluate approach • Adjust approach • Collect lessons learnt e. XPERT Best practices Bulgari a

r e m m u S l o o ch S Big Bang phase IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a Prerequisites • About 2 months length • The team has worked together on previous projects • The team never used XP or PSP before • The team had 1 week e. XPERT training • The customer did not have any XP experience e. XPERT Best practices

r e m m u S l o o ch S Big Bang phase - 2 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a all or nothing • Introduce all practices together – good or bad? • Default practices • Important practices • Sequential application – what the sequence should be? e. XPERT Best practices

r e m m u S l o o ch S Big Bang phase - 3 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J How to make the team do it? • Developers – right, wrong or just afraid? • Easily adopted practices • Not so easily adopted practices • Hard practices • Total control over the application of the practices e. XPERT Best practices Bulgari a

r e m m u S l o o ch S Big Bang phase - 4 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Initial evaluation and adjustment • Criteria for evaluation • Practices that needed immediate adjustment – Customer on-site – Planning game – Test first • Enough fooling about e. XPERT Best practices Bulgari a

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o Sch Sofia ly 2003 2 Ju une 30 J Productive Phase Get the best of it • Benefit from the strong sides – Identify and emphasize on the key practices – Minimize documentation • Minimize the effect of the weak sides – Adjust practices that imply big overhead, but preserve their benefits e. XPERT Best practices Bulgari a

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o Sch Productive Phase - 2 Key Practices • Which are they? Small Releases Simple Design Refactoring e. XPERT Best practices Sofia ly 2003 2 Ju une 30 J Bulgari a

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o Sch Sofia ly 2003 2 Ju une 30 J Bulgari a Productive Phase - 3 Adjust practices • Which practices? – Non-XP Customer – Planning game and Customer On-site – Unit testing – Pair programming – PSP PROBE method e. XPERT Best practices

r e m m u S l o o ch S Evaluation Phase IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a Collect quantitative measures • Why measure? – e. XPERT project requires it – PSP requires it – Increases developers discipline – Provides constant status feedback to management – Shows the practices that need adjustment e. XPERT Best practices

r e m m u S l o o ch S Evaluation Phase - 2 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Collect quantitative measures • What to measure? – Code metrics – LOC, NOM, … – Effort spent on tasks, processes, … – Costs – PSP PROBE method – effort per class, LOC, … – Defects statistics – new, fixed, closed, … – Unit tests – number of tests, failures, … e. XPERT Best practices Bulgari a

r e m m u S l o o ch S Evaluation Phase - 3 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a Collect quantitative measures • How to measure? – Automatic tools - code metrics, unit tests, bugs… – Manual measurements – effort, PROBE, costs, … • How often? – Daily – effort on tasks, PROBE, … – Weekly e. XPERT Best practices

r e m m u S l o o ch S Evaluation Phase - 4 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Main results • No schedule deviation • No costs deviation • Low requirements management risk e. XPERT Best practices Sofia ly 2003 2 Ju une 30 J Bulgari a

r e m m u S l o o ch S Evaluation Phase - 5 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert une 30 J Main Results Defect rates BEFORE NOW e. XPERT Best practices Sofia ly 2003 2 Ju Bulgari a

r e m m u S l o o ch S Evaluation Phase - 6 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Main Results Efforts per process e. XPERT Best practices Sofia ly 2003 2 Ju une 30 J Bulgari a

r e m m u S l o o ch S Evaluation Phase - 7 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Main Results Unit tests e. XPERT Best practices Sofia ly 2003 2 Ju une 30 J Bulgari a

r e m m u S l o o ch S Evaluation Phase - 8 IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Sofia ly 2003 2 Ju une 30 J Bulgari a Main Results PSP PROBE method Calculated hours per class type and size very small medium large very large simple UI classes 0, 3 0, 8 2, 0 4, 8 11, 5 property UI panel classes 1, 0 2, 2 4, 7 10, 3 22, 5 multi UI panel components 1, 4 4, 3 13, 2 40, 9 126, 6 model / data classes 1, 7 3, 0 5, 2 9, 1 15, 8 business logic classes 3, 1 5, 2 8, 9 15, 2 25, 8 utils/enum/interfaces/events/delegates 0, 6 1, 1 2, 1 4, 0 7, 5 e. XPERT Best practices

r e m m u S l o o ch S Nemetschek Pilot Project Sofia ly 2003 2 Ju une 30 J IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Tools used • • MS Visual Studio. NET – development IDE MS Visual Source Safe – version tracking Microsoft Project – effort and cost tracking MS Excel, MS Word Bugzilla – defect tracking NUnit – unit testing C# Refactory – code refactoring e. XPERT Best practices Bulgari a

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o h Sc. NUnit e. XPERT Best practices Sofia ly 2003 2 Ju une 30 J Bulgari a

r e m m u S l o o ch S C# Refactory IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert e. XPERT Best practices Sofia ly 2003 2 Ju une 30 J Bulgari a

r e m m u S l o o ch S Nemetschek Pilot Project Sofia ly 2003 2 Ju une 30 J IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert Lessons learned • • • CR management can be easy Design can be simple Coding can be the biggest part of a project Schedule can be kept Unit testing improves quality Estimation can be more precise, but can never be exact • e. XPERT is agile, lightweight and flexible • Documentation can be minimized e. XPERT Best practices Bulgari a

IST-2001 -34488 e. XPERT: Best Practice on e-Project Development http: //www. esi. es/Expert r e m m u S l o o Sch Sofia ly 2003 2 Ju une 30 J For more information: bangelov@nemetschek. bg expert@fmi. uni-sofia. bg http: //www. esi. es/Expert http: //www-it. fmi. uni-sofia. bg/Expert e. XPERT Best practices Bulgari a