I often use a makefile rule to generate "foo.exe" for each "foo.nim"
%.exe: %.nim
nim c -o:$@ $<
Run
To me the real problem is that "nim c" alters source directories by default. As
a general rule, it's helpful if a build system builds into the
current-working-directory, so you can simply delete the entire directory later.
mkdir build
cd build
nim c -r ../test/foo.nim
Run
Maybe it doesn't work that way because it would require unique names? Not sure.
As mentioned here, `nim r foo.nim` uses the cache-dir and avoids recompilation
on subsequent calls, which helps. But I sometimes have to define new
cache-directories. I work in an enterprise environment, not on my own private
laptop, so a shared cache-directory is problematic. That's not a complaint;
just something to beware.
