We need to improve the comments explaining the main ideas / insights of the solution

Currently, the bridged interactive grader will raise a CompileError if the interactor fails to compile. This is incorrect, and should be an InternalError instead.

Additionally, check_result will swallow any IEs if the submission's result_flag is non-zero.
This can lead to the case where the interactor crashes, and the submission reports an RTE/IR due to being unable to read from stdin.

It would be great to add a subset of the now only jury available stats also to the team and public pages.

I'm thinking of the submission graph over time (but all rejected verdicts bundled into one, of course nothing after the freeze) and the total number of attempts/corrects.

As you can see here, half of download_history.py is not covered by unittests.
At least, I want to test DownloadHistory.get fucntion.

We can test this module by 3 steps.

• Clear utils.cache_dir / 'download-history.jsonl'.
• Execute oj d, or directly instanciate DownloadHistory and

There are two sets of refactoring that can be performed on our codebase to make them more efficient and utilize the full potential offered by Qt Framework. See below:

1. There is no need to explicitly call delete on widgets that are allocated on Heap as long as in their constructor we have specified the parent widget for it. Consider this example below:

class Widget: public QObject