(src: Course: ADSD - Fallacies of Distributed Computing)

role: Business Analist

./resources/model-a-better-development-process-udi-dahan.resources/11460613-distributed-system-design-slides.1.jpg

role: Architect

./resources/model-a-better-development-process-udi-dahan.resources/11460613-distributed-system-design-slides.2.jpg

role: Project Manager

./resources/model-a-better-development-process-udi-dahan.resources/11460613-distributed-system-design-slides.3.jpg

Estimates Given a - well-formed team (worked together) - with required skills (has done such things before) - of size S

./resources/model-a-better-development-process-udi-dahan.resources/11460613-distributed-system-design-slides.jpg


I like how estimates formulated this way, drive Proof-of-Concepts with the sole focus of getting better estimates. This then lowers risk & variability by shifting that to the beginning of the project. Kinda the same principle as Technique: Walking Skeleton for incremental development where you put the path-to-prod at the beginning of a project, because it’s often the most variable thing.