Version 3.0.0

fantasy-land/traverse now takes a type representative rather than an of function (#220). Additionally, the order of the method's parameters has changed.

Version 1.0.1

Minor update that includes the following:

• Fixed the type signature for chainRec
• Typo's
• Bower registry

Version 1.0.0

Fantasy-Land version 1.0.0

Breaking changes

• ap arguments have now been flipped to keep in line with map and chain #145
• Specification now requires prefixed method names #146
• Traversable algebra now requires traverse method instead of sequence #135

Other improvements

• chainRec type has been added #152
• Fantasy-Land now builds against TravisCI #161
• Various typo fixes (if you see anymore PR's are more than welcome)
• Type signatures are now in the spec to hopefully make some of the spec clearer #147

Thanks goes to: @rpominov @davidchambers @safareli @scott-christopher @joneshf and many more.

Version 0.3.0

v0.3.0

Version 0.3.0

Version 0.2.1

Contains:

• Constant names for methods
• Addition of the laws for the specification

Version 0.2.0

v0.2.0

Version 0.2.0

Version 0.1.0

v0.1.0

Version 0.1.0

Version 2.2.0

Extend now depends on Functor (#213). Although the type-class diagram implied this, the textual component of the specification stated that Comonad rather than Extend depended on Functor.

The third Comonad law has been removed (#213).

Version 2.1.0

New type classes: Alt, Plus, and Alternative (#197).

Breaking changes to fix problems with fantasy-land/laws (#193). If depending on fantasy-land/laws one should lock major and minor version numbers (e.g. 2.1.x). The specification itself follows semantic versioning, but minor releases may continue to include breaking changes to fantasy-land/laws.

Version 2.0.0

This release introduces the concept of type representatives (#180). fantasy-land/empty, fantasy-land/of, and fantasy-land/chainRec must now be defined on type representatives. It's no longer necessary to look for definitions in two places – expressions such as typeof x['fantasy-land/empty'] == 'function' ? x : x.constructor are no longer required when defining polymorphic functions!

The type-class diagram in the readme now includes all the type classes defined in the specification (#192).