Permalink
Commits on Sep 8, 2010
  1. Fix typo in changelog

    indirect committed Sep 8, 2010
Commits on Sep 7, 2010
Commits on Sep 4, 2010
  1. Failing spec for #645, #652

    indirect committed Sep 4, 2010
  2. Minor improvements to generated gemspec.

    myronmarston committed Sep 3, 2010
    - Use ~> rather than >= for bundler version dependency, as per wycats' advice on his blog.
    - Simpler way to get executable files.
    - Added test files by interrogating git for test, spec and features files.
    - Test presence of gemspec file in newgem_spec.
Commits on Sep 3, 2010
Commits on Sep 1, 2010
  1. Fixed bundler/bundler#586

    spraints committed with wycats Aug 31, 2010
Commits on Aug 30, 2010
  1. Fix GemHelper#install_gem so that it raises an error rather than just…

    myronmarston committed Aug 30, 2010
    … printing it.
    
    The other gem helper methods raise errors rather than just printing them, so this aligns #install_gem with their behavior.
  2. Allow RSpec filtering.

    myronmarston committed Aug 27, 2010
Commits on Aug 29, 2010
  1. Merge branch '1-0-stable'

    indirect committed Aug 29, 2010
  2. Bump bundler to 1.0.0

    wycats committed Aug 29, 2010
  3. Bump bundler to 1.0.0

    wycats committed Aug 29, 2010
  4. Update CHANGELOG for 1.0.0

    wycats committed Aug 29, 2010
  5. Update CHANGELOG for 1.0.0

    wycats committed Aug 29, 2010
  6. 1.8 doesn't support :mode

    wycats committed Aug 29, 2010
  7. 1.8 doesn't support :mode

    wycats committed Aug 29, 2010
  8. Persist the new rubygem sources to the lockfile

    Carl Lerche committed with wycats Aug 29, 2010
  9. Persist the new rubygem sources to the lockfile

    Carl Lerche committed Aug 29, 2010
  10. Raise an error if the git push fails.

    myronmarston committed with wycats Aug 29, 2010
    We use the --quiet option so that we only get output for an error, and we assume any stderr output means the push failed.
    
    This closes #627.
  11. Pass the err buffer on to the block rather than re-reading stderr.

    myronmarston committed with wycats Aug 29, 2010
    This is necessary so that the code that checks stderr for error output can actually check the output of stderr, rather than an empty string.
    
    This closes #626.
  12. Raise an error if the git push fails.

    myronmarston committed with wycats Aug 29, 2010
    We use the --quiet option so that we only get output for an error, and we assume any stderr output means the push failed.
    
    This closes #627.
  13. Pass the err buffer on to the block rather than re-reading stderr.

    myronmarston committed with wycats Aug 29, 2010
    This is necessary so that the code that checks stderr for error output can actually check the output of stderr, rather than an empty string.
    
    This closes #626.
  14. Sanely add the "fake" version to the git source when the requirement …

    Carl Lerche committed Aug 29, 2010
    …is prefixed w/ equals
  15. Fix the logic that detects if the local clone of a git repository is …

    Carl Lerche committed Aug 29, 2010
    …up to date.