cat in create_task_makefile was writing different line endings depending on whether it was run from a Mac or a Windows machine, so I switched to readr::write_lines. (I've done this in other functions in the past; just missed this one.)
Current mismatches between machines will need to be rectified by having the windows users rebuild their task remake files with scmake(..., force=TRUE). My hope+intent is that there will be absolutely no changes to the .yml task files produced by Macs, but we should be vigilant for a while because there might be some. I had to remove a final newline character to get it to stay the same on the file I used to test this PR. I haven't tested on a task makefile that lacks a combiner function, which is the case I think most likely to break. (I actually think it won't break, but it still seems like the weak point).
cat
increate_task_makefile
was writing different line endings depending on whether it was run from a Mac or a Windows machine, so I switched toreadr::write_lines
. (I've done this in other functions in the past; just missed this one.)Current mismatches between machines will need to be rectified by having the windows users rebuild their task remake files with
scmake(..., force=TRUE)
. My hope+intent is that there will be absolutely no changes to the .yml task files produced by Macs, but we should be vigilant for a while because there might be some. I had to remove a final newline character to get it to stay the same on the file I used to test this PR. I haven't tested on a task makefile that lacks a combiner function, which is the case I think most likely to break. (I actually think it won't break, but it still seems like the weak point).