go-for-data-processing

### Go - for dataprocessing

why you wouldnt use go.
performance is not an issue

dataprocessing
compiled vs dynamic

if it is rigid vs dynamic python 
(rigid is good, for forcing)

### datateam - tv4
dataplatform started in golang
but switch to apache beam (java) of pipeline of jobs

go is not for building the whole pipeline
- jobs that use golang
- json to docker images

go channel is not for data processing
go channel locks the signals
go channel should be used to 

---
### go data science, data processing

packages:
pachyderm
- (using docker images w. json as a)
- versions the dataset for different aspects

Cern uses Golang for data science
- Gonum

Summary
use a typed language for data pipelines where rigid is inplace.
Most software and file formats in the data space is built on java, scala, kotlin.

THE cons of golang
is that it is hard to get golang to talk to another golang instance because of lack of Generics