Comment by saqwerUA on "I am Scott Hanselman. Ask me anything."

CommentI am Scott Hanselman. Ask me anything.

saqwerUA

Jan 4, 2018

Scott, hello from Ukraine! In my opinion, stack in Java is highly ready for set up BigData high load architecture (I mean Hadoop, Spark, Kafka, ElasticSearch and so on). DataScience's stack is mostly associated with Python and R now. When some projects need functional development - most of all they are looking for Scala, Haskell, Erlang and even Python engineers (but not F#). Could you explain, where is .NET doing better (except syntax)? How to compete with that tools?

Scott Hanselman

Open Source .NET/ASP.NET/C#/F#

Jan 4, 2018

I don't think .NET needs to compete. I think .NET needs to work well in a hybrid solution. Use the right tool for the job, no need to make a screwdriver into a hammer.

saqwerUA

Jan 4, 2018

As I know integration with Hadoop, Spark, Kafka, ElasticSearch and any tool which uses JVM is not a piece of cake. The same situation with Python (especially with projects and tools, which uses 2.7 version) to port someone's code to IronPython and back (if I want to share my changes with other people) is not a piece of cake as well. I'm not talking about Scala, Haskell, and Erlang. Sometimes the only option to integrate with them - design endpoint(-s) with declared API (which could be also a head pain as well). If you were me how would you try to resolve integration issues for each case?

Search Hashnode