Developers are constantly faced with a flood of new tools, creating a bias that newer technology is always the better choice.
Evaluating new tools is often clouded by cognitive biases like the sunk cost fallacy, Maslow's hammer, and the tractor factor.
New technologies carry a significantly larger magnitude of "unknown unknowns," increasing project risk compared to established tools.
Older, "boring" tools are often more reliable and cost-effective because their capabilities and failure modes are well understood.
A practical example demonstrates how to create a simple, fast, and effective browser automation test using the Selenium library in Java.
Using Maven simplifies project setup by managing dependencies like Selenium and JUnit and providing a standard build lifecycle.
A simple Bash script can be used to parameterize test runs, allowing the same Maven command to execute tests across different browsers.
A Makefile provides a simple, technology-agnostic entry point like `make test` to run underlying commands, standardizing workflows across projects.
AI models often provide more reliable and less hallucinatory results for older technologies due to the vast amount of historical training data available.
Be aware of personal biases, consider the value of familiarity, and decouple the idea of progress from simply adopting new technology.
Jobs that call for the skills explored in this talk.