I've been trying to port Haskell to the Open Source HPCC Systems platform (http://hpccsystems.com/) over the last couple of months, but been swamped by work and personal stuff, so I didn't move too far. In case you don't know much about it, this is a Big Data platform originally designed and built by LexisNexis (Reed Elsevier) about 15 years ago and released under an Apache 2.0 license a couple of years ago. This platform continues to be actively developed and extended and leverages clusters of nodes with share nothing architectures and local storage, over IP based networks.

It presents several nice synergies with Haskell: the platform is coded in C++ (as opposed to Java) and has a high level dataflow programming language called ECL, which is compiled to C++, and already allows for embedding C++, Java, Python, Javascript and R, so adding Haskell using any of those existing examples should be fairly trivial and could be probably done in just a few hours worth of work (see python, for example: https://github.com/hpcc-systems/HPCC-Platform/blob/master/plugins/pyembed/pyembed.cpp). Since ECL advocates non-strictness through purity too, with pure Haskell code the ECL compiler should be fairly free to distribute and parallelize Haskell code quite efficiently across the entire cluster. Another interesting aspect of the HPCC Systems platform is that it has a programmable distributed real-time data retrieval component called Roxie (programmable in ECL too), so this same work would also give Haskellers a distributed system for real-time large scale data delivery.

I'm posting this response here, with the hope that someone may be willing to help out with this integration effort too.

Kind regards,

Flavio