Samuel Olusola
May 2, 2024 ⋅ 11 min read

Understanding the dependency inversion principle in TypeScript

Samuel Olusola Software engineer (JS stack, GoLang incoming…) and student of computer science at the University of Lagos.

Recent posts:

Decoupling Monoliths Into Microservices With Feature Flags

Decoupling monoliths into microservices with feature flags

Explore how to effectively break down a monolithic application into microservices using feature flags and Flagsmith.

Kayode Adeniyi
Jul 25, 2024 ⋅ 10 min read
Lots of multi-colored blue and purplish rectangles.

Animating dialog and popover elements with CSS @starting-style

Native dialog and popover elements have their own well-defined roles in modern-day frontend web development. Dialog elements are known to […]

Rahul Chhodde
Jul 24, 2024 ⋅ 10 min read
Using Llama Index To Add Personal Data To Large Language Models

Using LlamaIndex to add personal data to LLMs

LlamaIndex provides tools for ingesting, processing, and implementing complex query workflows that combine data access with LLM prompting.

Ukeje Goodness
Jul 23, 2024 ⋅ 5 min read
JavaScript logo on top of violet background

Exploring essential DOM methods for frontend development

Learn four groups of DOM methods and their uses to create responsive and dynamic webpages. A helpful DOM reference table is also included.

Chimezie Innocent
Jul 23, 2024 ⋅ 12 min read
View all posts

6 Replies to "Understanding the dependency inversion principle in TypeScript"

  1. Hi, really nice article! A couple of typos in the code examples. You’re writing log.info instead of log.error when an exception occurs. Cheers!

    1. Thanks for the tip — would you mind pointing out the specific code blocks where the typos occur?

  2. A couple of problems with this principle. High level and low level is vaguely defined. If you apply this to the highest levels, this works fine. But the lower you go, the more this will feel the effects of an extra pointer to resolve or an extra function call. So, make sure that in your language, this results in, as much as possible, zero cost abstractions. Interfaces and Traits are typically fine, but watch out with proxies, abstract classes or any form of wrapper constructs.

Leave a Reply