Every time I hit a distributed systems problem between services in a service oriented architecture, I'm reminded how boring the average network failure problem is when you're just trying to write biz logic.

At a startup you don't have enough infra to solve the problem generically, so it's always busy work.

helping others build with LLMs and ML | YC S22, CMU CS + LTI | I write technical posts for ML builders at http://dbmikus.com

And my above cast looks like something you would barf up on an enterprise coding patterns blog.

At least if you code just in Solidity you don't need to worry about this because it's just big computer and the protocol engineers handle all the byzantine fault tolerance problems.