- some error occurred where the program never finishes and uses 100% cpu (maybe infinite while loop?)  so we capped the while loops at some value (100)

- writing to the stderror is now received (after every test line, only 1 line is read if there is any content in stderr)
- if the test is empty and the program outputs something then the test will fail

- empty tests should not be handled correctly (e.g. when the program waits for input then the timeoutis hit)
- if the user program throws an error the error is now written to the protocol

- when compiling 2 files the and no the main file had an error then the path is now removed (./ or .\)
- added to docs: * matches even if the program does not output anything
- expected output is now added as a new line in the conversation protocol

matches the new test runner/server interface