Changes in 0.1.0

FIT file reading is now handled by the fitdc package rather than compiled code from the FIT SDK, which threw up some licensing issues.
SRM file reading is now handled by pure R code, which removes the former python dependency and the readr suggest.
uniform_sampling is gone, replaced by resample, which is a much more explicit solution. Inspired by the resample method of the pandas DataFrame.
The clean_bikedata structure now includes a lap column.

Changes in 0.0.4

Code is now compatible with the latest dplyr release (>= v0.5.0). Most notably, it is now expected that contains is an exported function from dplyr.
magrittr is now an import. This was an implicit dependency anyway, as it is imported by dplyr. The pipe operator is now imported from magrittr rather than (convolutely) from dplyr.
This file now uses markdown rather than just plain text.

BUG FIX: A source file from the FIT SDK assumed a C++11-compatible compiler. Hence, C++11 is now listed in the package description file as a system requirement.
Cleaned bike data now has a "start_time" attribute where possible. This is facilitated by the fact that all importing functions now consistently try to append a "timestamp.posix" column.
Post-import processing functions (clean_bikedata and uniform_sampling) now better preserve the attributes of their inputs.

+ First official version; released on CRAN.