[PATCH conductor 0/2] Upstream cucumber fixes

Jason Guiditta jguiditt at redhat.com
Thu Jul 19 21:38:16 UTC 2012


On 05/07/12 16:45 -0400, Matt Wagner wrote:
>
>I've been trying to get Conductor running with rvm and Bundler, versus using system ruby. (I've been semi-inaccurately referring to this as "upstream" ruby, based on our desire to work upstream in pure ruby, without requiring Fedora.)
>
>I opened https://www.aeolusproject.org/redmine/issues/3447 to track getting cucumber tests to pass there.
>
>I ran these against ruby 1.8.7 patchlevel 357, using Bundler which got me 3.2-series Rails. With both of these in place, and various services running, I have only failure left (see task #3450 -- it seems to be reaching live Image Factory, and I'm not going to be able to fix this today, so I'm sending this out now.)
>
>Note that, ideally, this should be tested two ways -- one on "upstream" ruby with Bundler, and the other against system ruby to make sure I didn't break anything.
>
>We do want to get tests working with Ruby 1.9, and with all external services disabled, but that's not my focus with these particular patches.
>
>-- Matt

ACK and pushed.  Fwiw, this also fixed 7 failing specs for me, which
puts us down to 13 (with iwhd running), and 1 failing cuke under ruby
1.8.7 (with services running).  Ruby 1.9 had 3 failures, but ran in
3m13s vs ~15 minutes, wow! I also saw a speed increase on specs, but
forgot to track what they were.  Anyway, good work, hope to see others
pitch in on getting tests improved.

-j



More information about the aeolus-devel mailing list