
Security aspects
The evolution of the R language as a GNU distribution has been evolving over the past couple of decades as a client-based solution and, even before the dawn of the internet, all datasets have been securely stored on one's machine. With the spread of data over the internet, cloud services, and connected systems, as well as more and more present hybrid solution systems, the security aspect of reading, writing, analyzing, and storing data has never been a question. However, in the past ten years, more and more systems have been emerging, enterprise needs have changed, people roles have changed—but the core R engine and IDE has not yet taken a step in this direction. From this aspect, it seems as if R wants to remain a university developed and driven software. With R integration into SQL Server, Microsoft has addressed these issues and has not only created database roles, but also added services that make the connection to R and all computations securely, diminishing the possibilities of security vulnerabilities.