Wrong 0.4.0 released

2 views
Skip to first unread message

Alex Chaffee

unread,
Oct 4, 2010, 6:04:49 PM10/4/10
to wron...@googlegroups.com

== 0.4.0

* "require 'wrong'" and "include Wrong' now get everything; see README for details on how to get less than the whole enchilada
* array_diff replaced with string_comparison
* d
* use Sourcify (if available) to increase robustness
* new gem for jruby (gem install wrong-jruby) to fix gem dependency issues

== 0.3.3

* add JRuby support by removing ParseTree dependency
* minitest assertion counts work

== 0.3.2

* make Wrong work inside IRB
* stop embedding predicated and instead depend on its gem

== 0.3.1 - 2010/08/29

* remove "This is not true" messages, since they were more confusing than clarifying
* Test::Unit and Minitest pass assert/deny with no block up to the framework's assert method
* alias_assert and alias_deny (a truce in the language wars!)

== 0.3.0 - 2010/08/26

* color
* rename catch_raise to rescuing
* add capturing helper
* rspec adapter
* add a #close_to? predicate to Float and Fixnum (require "wrong/close_to")

== 0.2.0 - 2010/08/21

* Announced to the world
* Using Chunk

Alex Chaffee

unread,
Oct 4, 2010, 5:58:49 PM10/4/10
to wron...@googlegroups.com

thierry henrio

unread,
Oct 5, 2010, 7:06:56 PM10/5/10
to wron...@googlegroups.com
Hello!

On Mon, Oct 4, 2010 at 11:58 PM, Alex Chaffee <ale...@gmail.com> wrote:

== 0.4.0

* "require 'wrong'" and "include Wrong' now get everything; see README for details on how to get less than the whole enchilada
* array_diff replaced with string_comparison
* d
* use Sourcify (if available) to increase robustness
* new gem for jruby (gem install wrong-jruby) to fix gem dependency issues

Congrats!
I've done less than less, that is stumble on every bit of code I could slip on (days out)

There is now rspec2 tests, so I believe I should share some knowledge, if not code
- integrating rspec-2.0.0.beta.22 is straight

Rspec.configure do |config|
  config.include Wrong::Assert
end

And repository version is also valid, and I prefer it over rspec 'include' method
(I spiked it with a mongoid fork)

- integrating rspec-rails-2.0.0.beta.22 is not
at least, when looking hard at code, I found out that

require 'wrong'
RSpec::Rails::TestUnitAssertionAdapter::AssertionDelegate.send(:include, Wrong::Assert)

is a means of not having
     Failure/Error: assert {@cheddar.eater == @john}
     wrong number of arguments (0 for 1)

ouch!
I spiked it with an own rails project

Oh yes I could have asked rspec group for help :)

In a complete different topic, the following presentation http://www.slideshare.net/cperrone/outsanding-presentations-4351640 saved my yesterday :)
 
Cheers, Thierry
Reply all
Reply to author
Forward
0 new messages