Kauê Hunnicutt Bazilli

This seems to work in the current master: ``` ?- use_module(a). true. ?- test_a(X). X = a. ?- use_module(b). true. ?- test_a(X). X = a. ```

~I can reproduce this in the latest master in Linux.~ The "silent exit" is actually a segmentation fault. In Linux your shell actually reports this, so it's not really "silent",...

@flexoron, I think you are using the file from the other issue.

> or windows is hiding stderr? Segmentation faults aren't written to stderr as far as I know, it's just the shell that notices that a program has crashed and reports...

I actually can't reproduce this either in the _actual_ current master. (Note to self: Test things in the actual master instead of 3 commits back) It seems that commit 42d674950197882817e38d610de59f9d0cffa390...

Yeah, I don't know much about Windows, but there are lots of "Fault" in this event log you shared. Scryer is clearly crashing.

> So that is a problem in Scryer: importing a non-module file as a module is not OK, or at least worth a warning. Yes, when I was first learning...

I believe this issue is very important. It would be better to clarify this as early as possible to avoid headaches in the future. I believe that just putting somewhere...

Just noting that if the `initialization/1` directive is _after_ the `use_module/1` directive, this error doesn't happen: ```prolog %% Contents of initialize.pl :- use_module(library(debug)). :- initialization(run). run :- false. %% End...

It seems indeed to be the case that modules that define operators that are loaded after an `initialization/1` directive are the cause of this bug. ```prolog %% Contents of op_module.pl...