RSpec 1.3.0Edit
From the official release notes posted to the rspec-users mailing list:
rspec version 1.3.0 has been released!
* <http://rspec.info>
* <http://rubyforge.org/projects/rspec>
* <http://github.com/dchelimsky/rspec/wikis>
* <rspec-users at rubyforge.org>
* <rspec-devel at rubyforge.org>
Behaviour Driven Development for Ruby.
Changes:
### Version 1.3.0 / 2010-01-11
* enhancements
* capture ExpectationNotMet error by default in matcher DSL
* supports wrapping other expectations
* added match_unless_raises to matcher DSL
* supports easy wrapping of t/u assertions in matchers
* thanks to Chad Fowler and Glenn Vanderburg for the name
* add chain to matcher DSL (Joe Ferris - #935)
* see rdoc for Spec::Matchers
* and_yield provides configurable eval_context
* Eric Meyer & David Chelimsky
* CTRL-C actually stops execution! (Bryan Helmkamp - #911)
* make drb port configurable (Chris Flipse - #875)
* changed raise_error to raise_exception (#933)
* kept raise_error aliased, so this is completely
backwards compatible
* bug fixes
* don't define top-level context() method when running in IRB (#899)
* remove Ruby warning for uninitialized ivar (Bryan Helmkamp - #892)
* fully qualify reference to Default (Alex Sharp - #895)
* use runtime version of Test::Unit::VERSION in incompatibility message
(Ryan Bigg - #916)
* quote paths in RSpec's own specs so people running in paths with
spaces in them can achieve green too (Ryan Bigg - #917)
* ensure drb specs run when EADDRNOTAVAIL (Michael Klett - #881)
* throw_symbol matcher correctly bubbles errors up (#918)
* make Rakefile work without Cucumber installed (#919 - devrandom)
* escape . in regexp (#925)
* align be_true and be_false with Ruby's conditional semantics (#931)
* print duplicate nested descriptions (when appropriate) (#936)