Hmm, I'm seeing some sporadic test failures that I don't think are related to my work (though they could be.

===============================================================================
Failure:
  Expected <= "2038-01-01 00:00:00 -0800", but got 2038-01-01 20:03:45 -0800.
  <false> is not true.
test_between_with_time_parameters(TestFakerTime)
/Users/marklehman/Desktop/Code/faker/test/test_faker_time.rb:18:in `block in test_between_with_time_parameters'
/Users/marklehman/Desktop/Code/faker/test/test_faker_time.rb:15:in `times'
/Users/marklehman/Desktop/Code/faker/test/test_faker_time.rb:15:in `test_between_with_time_parameters'
     12:     from = Time.at(0)
     13:     to   = Time.at(2145945600)
     14: 
  => 15:     100.times do
     16:       random_time = @tester.between(from: from, to: to)
     17:       assert random_time >= from, "Expected >= \"#{from}\", but got #{random_time}"
     18:       assert random_time <= to  , "Expected <= \"#{to}\", but got #{random_time}"
===============================================================================
===============================================================================
Error: test_vin(TestFakerVehicle)
  NoMethodError: undefined method `match' for 5:Fixnum
/Users/marklehman/Desktop/Code/faker/lib/faker.rb:93:in `fetch'
/Users/marklehman/Desktop/Code/faker/lib/faker/vehicle.rb:15:in `vin'
/Users/marklehman/Desktop/Code/faker/test/test_faker_vehicle.rb:12:in `test_vin'
      9:   end
     10: 
     11:   def test_vin
  => 12:     vin = @tester.vin
     13:     checksum = vin_checksum(vin)
     14: 
     15:     assert vin.length == 17
===============================================================================

The test_vin one in particular is odd, because I didn't touch that file or test, which is what makes me think these sporadic failures are on master. However, I ran tests on master about a dozen times and didn't get these failures, so I'm at a loss. Any thoughts?

Interesting, travis is failing for ruby 2.0 because it doesn't seem to be respecting the kwargs in one of the files. Not sure what's going on there, I'll get ruby2.0.0 running locally and dive into it this evening.

I guess this was my lack of knowledge of the keyword arguments feature evolution in ruby. Apparently in 2.0 it raises a syntax error for method definitions like def some_method(first:, second:) such that you have to explicitly give it a default def some_method(first: nil, second: nil). Ruby >= 2.1 has no such problem with the first syntax.

I have updated the few method definitions that were failing with explicit defaults. I have made those defaults raise(ArgumentError) so that the user gets a uniform error for any faker method that they try to invoke with missing required arguments. Let me know if you don't think that's sensible.

I can write a few tests for that behavior as well.

Tests added, and build is green. Let me know what you think @stympy.

I added to the 2.0 milestone since these are BC-breaking changes.

Since we're migrating the objects to namespaces #1318, we'll change the arguments to keywords while adding the namespaces. I think it'd be a better idea to concentrate all the changes in this transition instead of doing things separately. Thanks for opening this PR and suggesting this change.

@vbrazo That makes sense, although looking through some of the comments and PRs linked on #1318 I am not seeing keyword arguments implemented, or even any mention of that they should be. We'll definitely want to make sure contributors are aware of that plan so they can implement their PRs accordingly.

@supremebeing7 you're talking about the lorem PR? I'm still working on this PR, and I'll include the keyword arguments :) I'll make sure to ask contributors to add the keyword arguments.

I just checked our master branch out and we only have a few namespaces and their methods don't have parameters.

Sounds good. Thanks for all your hard work.