Bindle Blog

In response to this experiment from a couple of years ago.

I often wonder about when it’s appropriate to use helpers or partials in Rails. There’s the ideological question (answered well in this post).

However, there’s also a performance question. You can’t help but feel that helpers are better when you see things like this in your logs:

Rendered selections/_list_item.html.erb (11.5ms)
Rendered selections/_frame.html.erb (17.0ms)
Rendered selections/_list_item.html.erb (0.3ms)
Rendered selections/_frame.html.erb (2.2ms)
Rendered selections/_list_item.html.erb (0.2ms)
Rendered selections/_frame.html.erb (2.0ms)
...and so on when you are rendering a lot of partials

However, this is development. Things are different on production. For instance if you follow Ben’s example and run tests on development, you might see (here’s the code):

ab -n 100 -c 5 -r http://helpers-vs-partials.dev/test/#{METHOD}

When `METHOD` is (result is req/s, higher is better):

partial: 2.20
helper: 5.93
partial_collection: 28.52

Which might lead you to make some kind of conclusions (probably quite similar to the ones Ben reached).

But, if you push the app up to a production server (heroku makes this super easy.. thanks guys!). You might see something more like:

partial:0.91
helper: 0.98
partial_collection: 1.01

Certainly not worth worrying about. QED.