Hacker News Reader

Comment by realrocker | Hacker News Reader

I have been researching other GO web frameworks last two days and this one is undeniably the best looking yet. But, I really wished they had integrated an ORM from the start. Also, I wish could know "why" they chose play framework as an architecture reference.

View on HN

I am definitely looking to develop a good story for ORM. However, I don't think it has to be in the very core of the framework -- allowing people to plug a different one (or not use one at all) is important.

I chose Play! (1.x) because it is by far the most productive development environment I have ever used (especially for boring business admin pages). (Out of Django, JSP, Lift, GSE/GXP, Jetty, etc).

Gorilla and pat (by the author of Sinatra) are probably the web "frameworks" I would recommend if you want to stick with minimalist idiomatic Go style.

http://code.google.com/p/gorilla/

https://github.com/bmizerany/pat

And just say no to ORM.

Disclaimer: I contributed to the original Gorilla design, but all the credit should go to moraes (and everyone else who was contributed to the project).

A lot of Go programmers, and I think I'd like to include myself here, are going to want to be able to plug in their own persistence backend.

For my own part, a lot of my work has some connection to a scaling problem of some kind. A mandatory/single-database-type ORM is totally out of the question and largely why I had to abandon Django for Flask in my Python work. I might at some point migrate to Pyramid for Python, but I haven't evaluated the state of their persistence story. I was mostly wooed by their testing strategy (it's drop dead sexy).

Reply on news.ycombinator.com