sre990/ske-pi

Data parallel and stream parallel skeletons implemented in erlang

ske-pi's Issues

Documentation file

A more in-depth look at all functionalities + how to use this library would be neat! :)

Usage/helper function

Adding, for each module, a helper function printing possible usage options.

Better error handling

As of right now, these modules have been developed with little to no concern towards error handling. It would be better to improve error handling and use the LOG macro.

Binaries instead of lists

The Erlang concurrency model is shared-nothing, meaning that all data must be copied
between processes and this introduces some overhead. Overhead may be minimised by using binaries instead of lists, because binaries are passed by reference.

Gmapred data structures comparison

Right now, gmapred is using erlang's dict module. Try to look for a more efficient data structure and compare performance.

Gmapred example

Adding a gmapred example test. Perhaps a simple "grep" operation?

Escript examples

As of right now, examples are merely modules used only via the erl emulator. It would be better to turn them into scripts.

Static typing

type-checking done statically, using dialyzer and adding type annotations