Go's standard library is excellent. I've found Gorilla Mux to nicely augment net/http and not feel like I'm wrapping magic around idiomatic Go. The only other 3rd party package I routinely use is sqlx.

http://www.gorillatoolkit.org/pkg/mux

https://github.com/jmoiron/sqlx

Give Squalor (https://github.com/square/squalor) a shot.

You might find https://github.com/relops/sqlc really nice from an API design point of view (inspird by jOOQ, a Java sql library).

Sqlc is no longer under development, but you can easily use it as a reference to adapt other libraries to have a similar API.

I literally found out about sqlx a few moments before coming here. I was stringing together ~10 columns by hand, and then when i fed them into the `row.Scan()` method, i just about rage quit. On Monday i'm going to try integrating sqlx and see how i like it.

Any other good sql libs i should know about? My coworkers mostly like sql by hand, so no magic ORMs preferably. But unmarshalling rows seems like an absolute requirement.. this is just anarchy without it.

Gorm. Give it a try. Light on magic. Can write by hand easily still. If you arent writing really complex queries and are just shuffling records around Gorm makes life a joy. It is no SQL alchemy but for most SQL users ORM/SRM makes life better. Anyway I think most systems using SQL in business logic end up implementing a relational mapper, just not as well as battle tested mappers. But some use cases do justify avoiding such tools.

Love mux, its super simple and still lets me use the standard library way of making web apps

I generally dislike the little things that beautify things. I don't like the decorators for routing in Flask, for example.

This is why I like CGI. I can make an effort to understand it.

Then you are absolutely going to hate the article I recently contributed to Gopher Academy's advent series, https://blog.gopheracademy.com/advent-2016/go-syntax-for-dsl... , about using Go to create terse, declarative DSLs for things like, say, building a webapp.