[rspec-users] [Cucumber, Webrat] Error backtraces shown as HTML in console

Matt Wynne matt at mattwynne.net
Thu Feb 5 07:35:13 EST 2009

On 5 Feb 2009, at 11:00, aslak hellesoy wrote:

> On Thu, Feb 5, 2009 at 10:38 AM, Matt Wynne <matt at mattwynne.net>  
> wrote:
>> Since upgrading to Cucumber I've noticed that the  
>> backtraces
>> appearing in the console output are of the form:
>>   Page load was not successful (Code: 500):
>> Followed by the dump of the rails error page. That page looks fine  
>> in a
>> browser, and it's helpful enough to see it when save_and_open_page  
>> does its
>> thing, but I'm pretty used to reading plain old stacktraces in the  
>> console -
>> the HTML output is very hard to read.
> Rails' HTML error page or Cucumber's HTML formatter? (The latter is
> very rudimentary in 0.1.99).

Rails' HTML error page. It comes up both in my browser (presumably  
save_and_open_page doing its thing) and the same HTML is dumped to the  

For example, this is coming out in the console:

     <p>Extracted source (around line <b>#1</b>):
       <pre><code>1: &lt;% if concert.any_media? %&gt;
       2:   &lt;% media_summary_images do |images| %&gt;
       3:     &lt;div id=&quot;media&quot;&gt;
       4:       &lt;h3&gt;Photos from previous concerts&lt;/h3&gt;

>> Is this a known issue, or a feature, or have I just got something  
>> configured
>> wrong?
> It's an unknown issue (bug). Try --backtrace.

Thanks for the suggestion. I tried that, it doesn't seem to make any  

> Please file a bug report.

OK, but I am really surprised nobody else is getting this - it feels  
more to me like I've got some configuration setting wrong, or a  
mismatch in Cucumber / Webrat versions - our setup has been doing a  
long while and doesn't always have the very bleedingest edge (I think  
we're still on webrat 4.0, for example) and there is sometimes crufty  
lying around that gets in the way.

Can anyone else confirm that with latest Cucumber and Webrat they  
*don't* get this behaviour?

Matt Wynne

More information about the rspec-users mailing list