Technically it does run on windows, you just can't build the entire dataset without adding the sharding code mentioned. Set divisor=100 in config.properties an

Technically it does run on windows, you just can't build the entire dataset without adding the sharding code mentioned. Set divisor=100 in config.properties and it will happily build an index over 1% of the dataset.