Hi all,<br>A bit of feedback I got from my talk was that people were still unsure of exactly how/where to use RubyProf in their code.<br><br>I didn't cover it much because there are some great examples of this on the ruby-prof homepage here: <a href="http://ruby-prof.rubyforge.org/">http://ruby-prof.rubyforge.org/</a><br>
It pretty much covers all cases.<br>One issue I did have with the examples was with RubyProf::Test - it is meant to run each test case ten times to give you better average results... but it doesn't re-run your setup method each time... which renderes it practically useless if you do anything significant in setup... and given we use shoulda - that means every test case :(<br>
<br>I hand-hacked up a solution which I cover in a blogpost here:<br><a href="http://rubyglasses.blogspot.com/2009/08/faking-startup-and-shutdown-with.html">http://rubyglasses.blogspot.com/2009/08/faking-startup-and-shutdown-with.html</a><br>
<br>It doesn't do multiple runs over the tests, but does allow you to run RubyProf over your whole test-suite.<br>Be aware that it'll generate a *huge* callgrind file that may take forever to a) print and b) load up in kcachegrind.<br>
<br>If there are any other questiosn - feel free to ask :)<br><br>Cheers,<br>Taryn<br>