golang-standards/project-layout VS service

cmd: here where we will leave the main.go that starts our app.

> https://github.com/golang-standards/project-layout

The name of the repo is really and intentionally misleading. rsc filled an issue there to point this out, but the repo maintainer just disabled issues altogether so now no one can see it.

Even when it would not have such parasitic name, many seasoned Go programmers, me included, consider the self-proclaimed "Standard Go project Layout" as the opposite of what is good and advisable.

Unfortunately, the name just works, so it is being recommended all over the Internet since its inception.

Your comment makes it look as if you're saying you hate the Go development team, but it seems that isn't the case.

I get a little of what you're saying, I wouldn't say I hate anyone, but I strongly dislike how a lot of projects are organized. I think a lot stems from https://github.com/golang-standards/project-layout , which pretended to be standard and was so (ab)used one of the creators opened an issue about it. If you look at the actual Go src, it's much, much cleaner.

So you don't get blind sided for self proclaimed "standards" that are not

I am writing a project that has two components: a CLI and a library. I've organised the project as follows (as per organisation instructions):

Familiarity with the standard Golang project structure, which you can find here.

I've been learning how to build web applications using different frameworks and languages for a while now, such as Laravel with its MVC architecture and Node.js following the 'Hapi.js Way'. As I'm trying to create a new portfolio project using Go, I found myself contemplating over the ideal project structure. I wanted something that not only aligns with the standard Go project layout, but also makes the code both easy to write and understand. That's when I stumbled upon the concept of Hexagonal Architecture, as showcased in Netflix's engineering blog. The idea of seamlessly swapping infrastructures with minimal code changes fascinated me, and I decided to implement it in my new project.

Ultimate Service was useful for me. None of the "backend" concepts were new, but you get to see how Bill would layout/design an API-based service. If you're experienced you'll notice the opinionated choices he makes, and I found myself saying "Nah, I'm not sure I'd do it like that". I appreciated its use of Kubernetes and KIND as I'd never played with them before. How he uses Docker to spin up a DB instance for tests is pretty cool. There's a lot of copy & paste as you code along with him (you copy from the "finished project" and paste into your work in progress). The full example project is online at https://github.com/ardanlabs/service. You won't write all that code, and this version is newer than the one I did, but it gives you an idea of what you might learn.

So what can you do insted? For testing databases, setup a docker instance for tests (e.g. like in https://github.com/ardanlabs/service), or start an embedded-postgres daemon (see https://github.com/fergusstrange/embedded-postgres). For communication with external APIs, just pass the http.Client (either in context.Context or as a field on the struct). Then in tests, you can override the http.Client.Transport func.

Take a look at https://github.com/ardanlabs/service from Bill Kennedy. You can probably simplify the structure a bit since your project is minimal, but that repo is gold.

Checkout https://github.com/ardanlabs/service for inporation. Tip: try to avoid creating a service package with all services, a domain package with all domain structs, etc.

Or go through this repo https://github.com/ardanlabs/service

I've seen this handler modification and wrapping pattern in Ardan Labs' service repository. https://github.com/ardanlabs/service/tree/master/foundation/web

https://github.com/ardanlabs/service This is something which I really like and has taken into account a lot of engineering decisions.