Reliable systems must behave well in the face of many failure scenarios. These failures can be at different levels, and sometimes we don't even know them in advance.
This unpredictable, unexpected, nature of failures ("essential" complexity of the problem domain) perpetually threatens to inject itself into our solution, as implementation-specific "incidental" complexity.
Choosing the correct abstractions, including judicious use of Clojure's macro system, can help achieve these apparently conflicting goals of reliability, scalability, and simplicity.
In the talk Kapil will cover:
Different failure scenarios and techniques to address them
How to decide when to choose which abstraction?
A FSM controlled by input and control signals
Using components to build systems
Verifying the systems using DSLs built with macros
General Clojure good practices that make development a lot easier
29 авг 2024
