NickCH-K/did

.cmd hangs on installing dependencies

Danferno opened this issue · 6 comments

afbeelding
There's a flashing underscore but nothing seems to be happening?

I closed the window and now I get a lot of empty .cmd windows opening, sometimes they show here() starts at C:/StataWD

I think maybe it was installing the packages but didn't flush text to the cmd window? At least something similar happened when I installed in R directly and then more output appeared when I clicked in the window. Maybe add a warning that this can take a while?

The empty .cmd windows are rcall doing its thing. It likes to open and close a lot of .cmd windows. Those are the flashing lights I warn about in documentation.

The hang is a bit concerning. Just happen to be working now on documentation updates about this.

Would there be a way of just running an R-script once instead of the separate calls? Or are you dependent on Rcall for that? (I'm not an R user)

It produces a lot of .cmd windows even with only one rcall call. I've done some things to slim down the number you get in most parts of the package. Although in the case of didsetup the checks for package installation need to be their own calls, so that increases things a bit.

Fixed (to the extent possible) with f4e2467