# Canoo webtest

**URL:** <https://www.ruby-forum.com/t/canoo-webtest/158538>\
**Category:** Rails\
**Created:** [January 24, 2009, 2:11pm UTC](https://www.ruby-forum.com/t/canoo-webtest/158538 "2009-01-24T14:11:39Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![bahaw](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/b/f19dbf/32.png) [@bahaw](https://www.ruby-forum.com/u/bahaw)\
**Post date:** [January 24, 2009, 2:11pm UTC](https://www.ruby-forum.com/t/canoo-webtest/158538/1 "2009-01-24T14:11:39Z")

</div>

has anyone used canoo webtest to test a rails application?  
appreciate the info.

---

<div class="post-metadata">

**Author:** ![bahaw](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/b/f19dbf/32.png) [@bahaw](https://www.ruby-forum.com/u/bahaw)\
**Post date:** [January 24, 2009, 5:00pm UTC](https://www.ruby-forum.com/t/canoo-webtest/158538/2 "2009-01-24T17:00:52Z")

</div>

Bahaw A. wrote:

> has anyone used canoo webtest to test a rails application?  
> appreciate the info.

Firstly, one ANT configuration file can be longer than an entire Ruby  
module. If  
you go that route, you will soon encounter this phenomenon:

[http://c2.com/cgi/wiki?ConfigurationHell](http://c2.com/cgi/wiki?ConfigurationHell)

Next, canoo hits a web server to test. This is bizarrely inefficient if  
most GUI  
testing should happen directly to the XHTML that a server would have  
served.  
Rails “functional” tests call actions in controllers, and returns their  
contents  
as strings for parse-testing. This provides 95% of the coverage needed.

The remaining coverage happens in live JavaScript. That is where we need  
in-browser testing, such as Watir, Selenium, or Firewatir.

Tests that run a webserver, but then don’t run a web browser, are the  
admission  
of defeat before you start. They neglect simply generating a page and  
not  
serving it.

–  
Phlip