version 1.0.2
- Imporved sanity checking on input data frames (Thanks to Brian Ripley for
  pointing this out)

version 1.0.1
- Updated compiled code to new compiler prototype warnings.

version 1.0.0
- Fix: for integer vectors, the output would in some cases depend
  on the order of the input. (Thanks to Tobias Rockel for reporting).

version 0.2.2
- Fix: for small numbers of records and high number of threads, the output
  would depend on 'nthreads'. (Thanks to Matt Austin for reporing). Gower
  now only paralellizes when more than 1000 records are presented.

version 0.2.1
- moved test framework to 'tinytest'

version 0.2.0
- gower_dist and gower_topn gain `weight` argument for weighted matching (thanks to David Turner)
- gower_dist and gower_topn gain `ignore_case` argument for automatic column matching.
- gower_dist now returns numeric(0) invisibly when there are no columns to compare.
- gower_topn now returns a list with empty matrices when there are no columns to compare.
- gower_topn now warns when n>nrow(y) and sets n=nrow(y)
- bugfix: comparing factors with characters would cause a crash (thanks to Max Kuhn)

version 0.1.2
- fixed valgrind warning
- registered native routines, as now recommended by CRAN

version 0.1.1
- gower_dist now issues warning when nr of rows on input do not divide (recycling)
- Code now depends on gcc version for compatability with Windows + R<=3.2.x
- Bugfix: bad range calculation caused faulty distance computation

version 0.1.0 
- First release