You should probably mark the question as answered and ask a new question. I think you'll get more readers who could potentially answer your new question and it is easier in the long term to have separate issues in separate questions.
I pretty much always start in a vanilla R session with nothing loaded and then read in datasets with functions like read.csv(). Since I do all my work in scripts (.R or .Rmd), it's pretty simple to run the already written line of code again. However, you can have R save and then load a .RData file into your workspace on start up. I can't tell you much about it since this is a feature I've never used it 
, but someone may come along to your new question and be able to give you more details.