3 ms·
looks like dask is python-only, so it's a nonstarter (loser) for already existing JVM code that runs on spark
by maximente 7y ago
looks like dask is python-only, so it's a nonstarter (loser) for already existing JVM code that runs on spark
- missosoup 7y agoSpark stacks inevitably end up with PySpark though. It's rework for people who already committed to Spark, sure. And for bigger projects that committed to Spark this change isn't justifiable. But for a greenfield project, choosing Spark is just silly today.