The common belief that architecture should be future-proof is a flawed conclusion drawn from the fact that architectural decisions are hard to change.
A real-world project failed because the team discovered their initial migration plan was flawed only after the project began, highlighting the danger of rigid upfront plans.
Architecture must be iterative because domain models evolve, requirements change, and developers continuously learn and improve their skills.
Instead of over-planning, start with domain prototyping to focus on business logic first and defer technology decisions to the last responsible moment.
While YAGNI was a useful counter to big upfront design, a better approach is to navigate immediate requirements while always keeping the long-term project goal in view.
Aiming for a future-proof architecture creates a reluctance to make necessary changes, leading to a messy system as new features are forced into an outdated structure.
Effective architecture is not about abstract technical diagrams but about creating a structure that clearly reflects the business domain and its processes.
Preparing for technology changes by creating technology-independent code is often unnecessary, as major technology shifts usually coincide with required changes to the business logic.
The key to long-term project success is not a future-proof plan but an adaptable architecture that is driven by the domain and accepts the limited lifespan of technologies.
Sopra Steria Custom Software Solutions GmbH
München, Germany
Finanz Informatik
Frankfurt am Main, Germany
24:52Hendrik Lösch
42:02Simon Lasselsberger
43:13Carola Lilienthal
26:15Gernot Starke & Benjamin Wolf
21:14Andrey Semenyuk
21:39Johannes Hansen
46:48Stefan Priebsch
28:11Simon Lasselsberger
.gif?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
PDV GmbH
Erfurt, Germany
esentri AG
Rocken AG
Burgdorf, Switzerland
ADVERGY GmbH
Köln, Germany
Wilhelm Layher GmbH & Co KG
Oberes Zabergäu GVV, Germany
