Bugfixes:

• Remove duplicate numbers import from the bower.json file to support publishing to Pursuit (#128 by @thomashoneyman)

Breaking changes:

• Migrate FFI to ES modules (#130 by @kl0tl and @JordanMartinez)

• Replaced polymorphic proxies with monomorphic Proxy (#132 by @JordanMartinez)

• Make frequency use NonEmptyArray (#131 by @JordanMartinez)

  Now oneOf and frequency both use NonEmptyArray rather than NonEmptyList.

Bugfixes:

• quickCheckPure and quickCheckPure' stack safety (#127)

Other improvements:

• Drop math dependency; update imports (#133 by @JordanMartinez)

New features:

• Export enum and randomSampleOne (#125 by @JordanMartinez)

Other improvements:

• Fix warning revealed by v0.14.1 PS release (#125 by @JordanMartinez)

Breaking changes:

• Added support for PureScript 0.14 and dropped support for all previous versions (#115)
• Fixed genericCoarbitrary so it no longer explicitly relates to the generated data (#99)
• Replaced NonEmpty Array and NonEmpty List with NonEmptyArray and NonEmptyList (#118)
• Replaced globals dependency with numbers (#120)

New features:

• Added randomSampleOne (#114)

Bugfixes:

• Allowed full Number range for chooseFloat (#97)

Other improvements:

• Dropped generics-rep dependency as its now included in prelude (#121)
• Replaced monomorphic proxies with Type.Proxy.Proxy and polymorphic variables (#116)
• Migrated CI to GitHub Actions and updated installation instructions to use Spago (#117)
• Added a changelog and pull request template (#122, #123)

Added some functions to give better reporting abilities for pure QuickCheck runs

• Move guide over from documentation repo (@anttih)
• Bump dependencies

Updated for PureScript 0.12

• Added quickCheckGen and some variants that specialize the properties to be Gen-based (for convenience when using MonadGen-constrained properties)
• Added instances for NonEmptyArray and NonEmptyString (@matthewleon)

• Fixed resize resetting seed (@matthewleon)

• Fixed behaviour of resize (@safareli)

Previously resize would alter the state of Gen forever, rather than it being local to the Gen that was passed into resize. This may alter the behaviour of some existing tests if they relied on the buggy behaviour.