4 ms·
I'm very excited to see work like this. There's no reason to accept MapReduce as the best type signature for structuring distributed, parallelizable algorithms.
by jganetsk 17y ago
I'm very excited to see work like this. There's no reason to accept MapReduce as the best type signature for structuring distributed, parallelizable algorithms. Most of the justification is that it "feels right", and "just works".
- sgk284 17y agoThanks jganetsk, words of encouragement are always appreciated.