The new Joins in data.tablevignette is great, thank you!

I think the updating by reference section could improve by:

1. provinding a simpler example with simple one-to-one match example. In the current example the beauty of the NO-COPY solutioun is a bit obscured by the complicated grouping ("last element selection", "tail", etc) used.
2. provide an efficient solution for a RIGH JOIN that minimizes copying. At least in my workflow, that is the main use case. I have a main DT, on which I performe several join operations a[DT,on=.(idA)], b[DT,on=.(idB)],.... I assume this is a super common usecase and would be great to discuss the canonical solution to it (even if the answer is that the x[i,on=, j =] syntax does not allow for updating i by reference)

In this SO question, I provided sample code for 1., with a simpler join example, and 2., with my current solution to avoid the expensive copy.

UPDATE1: the solution proposed in SO is super elegant and performant:

x = data.table(id = c(1:5,8), newvar1=c(LETTERS[1:5],'h'),newvar2=c(5:1,-2))
#In practice x would have more vars: newvar2, ..., newvarN 
i = data.table(id = 1:7,var1 = c('bla','ble','bli','blo','blu','blA','blS'),var2=7:1,var3=2*(7:1),var4=1)
cols <- setdiff(names(x), 'id')
i[, (cols) := x[.SD, on = "id", .SD, .SDcols = cols]]

This should definetivelly go on the tutorial I wish I had known this ealier. On my application ( i: 220m rows, 40ish columns. x: 130m and 5 cols), the performance gain was massive. This method is 3x faster and uses only 1/5th to 1/6th of the RAM increase relative to i <- x[i,on=.(id)]

[I could not find an issue specific to this vignette, just the general #944]