Tharun

@Akintola_steve Wow! This is really helpful. I have worked with different PSPs but never thought to handle it to this extent.

Bookmark this tweet right now. Not after reading. Now. At some point, you will build a payment system, work on one, or inherit one that is already broken. When something goes wrong, and it will, you will need exactly what I am about to show you. This is the most complete payment system breakdown I have ever written. Every flow. Every edge case. Every trap. One thread

@pulkit_mittal_ Will do that... Thank you

@_tharun_d Hey! Start with the basic ones like url shortener, pastebin,etc.

Most engineers find System Design difficult not because they can’t design systems, but because they try to memorize architectures instead of understanding first principles. Here are 3 core system design theorems / concepts that explain why systems behave the way they do. 1. CAP Theorem CAP says a distributed system can guarantee only two of: - Consistency - Availability - Partition Tolerance Once a network partition happens (and it will), you must choose: - CP: correct but may be unavailable - AP: available but eventually consistent ⸻ 2. ACID vs BASE Strong guarantees come at a cost. - ACID systems optimize for correctness - BASE systems optimize for availability and scale If you understand: - When strict consistency is necessary - When eventual consistency is acceptable You stop over-engineering databases and start designing practical systems. ⸻ 3. Idempotency & Exactly-Once Myth Retries are inevitable. Exactly-once is mostly a lie. If you understand: - Idempotent APIs - De-duplication keys - At-least-once delivery semantics You can design systems that survive: - Network failures - Client retries - Partial writes Reliable systems assume retries will happen and design for it.

@marcoieni I tried anyhow but then i needed some centralized error enum in my projects so i only use thiserror instead.

A simple way of handling errors in Rust using the anyhow crate