Randy Lai

For the record, R 3.5.1 works in my macOS.

I am using the R from CRAN though. I don't expect anything would break on R 3.5.1 and there is no reason to break. Btw, is there anything in `build.log`?

@jotwin Were you using X11 as the graphical device? Would you try using pdf device? ``` R""" pdf("a.pdf") print(qplot(data=data.frame(x=rnorm(100), y=rnorm(100)), x=x, y= y) + geom_smooth(method='loess')) dev.off() """ ```

Does it work without `geom_smotth`? (i.e., it creates the file `a.pdf`). How about running the same commands in R? Does it raise a segfault?

Please check if 72dc8a1 fixes your issue.

Try again 2815fa9

As far as I know, the current version of RCall depends on DataFrame.jl v0.21 but RDatasets.jl doesn't support it.

There is also a related issue when doing `lm`. (PS: please checkout current master for DataFrame conversion) ``` julia using RCall using RDatasets mtcars = dataset("datasets", "mtcars"); rprint(rcall(:lm, "Disp~MPG", data=mtcars))...

This may be a better way to resolve the lazy evaluation problem. It uses `R_mkEVPROMISE` to create a promise object first. ``` julia using RCall RCall.rgui_start() macro promise(x) sym =...

It is used to disabled [reference counting](http://developer.r-project.org/Refcnt.html) so that `gc()` will collect the corresponding memory. In default, R is using a different way to do reference counting. It is not...