Getting very close. Right now there are 3 errors reported by addons/drb, all other specs pass. That doesn't mean everything will work, but we're in good shape. Only problem is, RJB 1.1.6 will install on Windows but nowhere else (compilation issues), RJB 1.1.7 installs everywhere but Windows.
http://github.com/assaf/buildr/tree/ruby1.9 Assaf On Thu, Jun 18, 2009 at 12:29 PM, Assaf Arkin <ar...@intalio.com> wrote: > In lib/buildr/scala/tests.rb lines 112, 113: > ant.includes group_includes.join(" ") if group_includes > ant.excludes group_excludes.join(" ") if group_excludes > > The arguments should be name/value pairs, a string argument doesn't do much > good. Commenting out these two lines doesn't break tests_spec.rb. > > Assaf > > > > On Wed, Jun 17, 2009 at 8:37 AM, Assaf Arkin <ar...@intalio.com> wrote: > >> On Wed, Jun 17, 2009 at 8:28 AM, Daniel Spiewak <djspie...@gmail.com>wrote: >> >>> Turns out that I just hadn't updated the specs to add the $ back into the >>> expected lists. Specifications are defined as singleton objects, which >>> means that the test class which is running is actually '<specName>$'. I >>> used to trim off that trailing $, but that isn't really an option with >>> the >>> new spec runner. Ideas welcome here. >>> >>> Incidentally, the new spec runner does provide some very important >>> benefits >>> -- like being able to detect and run *any* bona fide specification >>> object, >>> not just the ones which lack companion classes. Also, this spec runner >>> makes it possible in future to run specifications defined as classes >>> (rather >>> than objects), something which is supported by Specs but not Buildr. So, >>> the solution isn't to just drop the new runner; I just have to figure out >>> a >>> way to make this work without exposing all those ugly $ characters to the >>> end-user. >>> >>> In the meantime, I've fixed the specs and committed the results. >>> >> >> Awsome. I'm working on getting all the specs to pass with Ruby 1.9, one >> set at a time, so I need specs to first pass with 1.8. >> >> Assaf >> >> >>> >>> Daniel >>> >>> On Tue, Jun 16, 2009 at 9:53 PM, Daniel Spiewak <djspie...@gmail.com> >>> wrote: >>> >>> > Those tests were passing for me yesterday, before the Rspec upgrade. >>> Well, >>> > *some* of them were passing. Most were throwing errors about a helper >>> > method. I'll take a look at it tomorrow to see if I can reproduce the >>> > failures. >>> > >>> > Daniel >>> > >>> > >>> > On Jun 16, 2009, at 8:33 PM, Alex Boisvert <boisv...@intalio.com> >>> wrote: >>> > >>> > Hi Daniel, >>> >> >>> >> Assaf and I noticed some failures in spec/scala/bdd_spec.rb today. Do >>> the >>> >> following specs fail for you? >>> >> >>> >> 1) >>> >> 'Buildr::Scala::Specs should include public classes extending >>> >> org.specs.Specification' FAILED >>> >> expected ["com.example.MySpecs$"] to include "com.example.MySpecs" >>> >> ./spec/scala/bdd_spec.rb:68: >>> >> >>> >> 2) >>> >> 'Buildr::Scala::Specs should include public classes extending >>> >> org.specs.Specification even with companion classes' FAILED >>> >> expected ["com.example.MySpecs$"] to include "com.example.MySpecs" >>> >> ./spec/scala/bdd_spec.rb:85: >>> >> >>> >> 3) >>> >> 'Buildr::Scala::Specs should report failed test names' FAILED >>> >> expected ["FailingSpecs$"] to include "FailingSpecs" >>> >> ./spec/scala/bdd_spec.rb:128: >>> >> >>> >> 4) >>> >> 'Buildr::Scala::Specs should compile and run specifications with >>> "Specs" >>> >> suffix' FAILED >>> >> expected ["HelloWorldSpecs$"] to include "HelloWorldSpecs" >>> >> ./spec/scala/bdd_spec.rb:146: >>> >> >>> >> 5) >>> >> 'Buildr::Scala::Specs should fail if specifications fail' FAILED >>> >> expected ["StringSpecs$"] to include "StringSpecs" >>> >> ./spec/scala/bdd_spec.rb:165: >>> >> >>> >> Finished in 14.103377 seconds >>> >> >>> >> 11 examples, 5 failures >>> >> >>> >> alex >>> >> >>> > >>> >> >> >
