Weeks Twelve and Thirteen EPF6

--- title: Weeks Twelve and Thirteen EPF6 tags: [EPF] --- # Work With No Visible Result Since the last update, which was less than two weeks ago, I spent time on first writing tests for my PR and second, trying to resolve the issue with the Ephemery network on Teku. Third, looked into failing the test of Besu written a long time ago. Then there's the review I got from Besu, which is progressing, and it's going to be the main thing on my plate for next week. I worked on tests for a while and realized I need to spend more time to find a test scenario that fabricates the functionality of the restart feature. This needs a confirmed and merged implementation so I won't spend lots of time on implementing tests for unmerged code on my PR. Therefore, I didn't spend more time on it. I should spend more time on this later when the implementation of Ephemery is merged, as the tests are tailored specifically to the implementation. Then I started working on the issue with Teku that it doesn't work with the network ephemery command and only works with the URL or a local config YAML file. Guess what? The next day I realized the problem no longer existed; Paul Harris resolved the issue. 🫡 The other thing I spent time on was old tests of Besu that had been written by team members, probably a long time ago. The problem is these tests react differently on different machines. Some of them don't pass on my machine, while on others they pass with no issue. The problem arises from the fact that on some machines it translates a named DNS into a numeric one. An outcome is that when the test is successful, it responds with some sort of `endpoint_not_valid` message due to the arbitrary `garbage.endpoint.com` address. But it succeeds because the underlying library method changes it to `garbage.endpoint.com/198.18.0.77`, whereas it should have failed. I spent time resolving the problem, but the code snippet breaks other tests. Long story short, the problem still exists; you can follow the conversation [here on Besu Discord.](https://discord.com/channels/905194001349627914/1412446473907474462) Working on tests; even implementing a feature on Besu needs running tests. Every test run takes more than 30 minutes on my machine and 18 minutes on a team member of Besu's, still too time-consuming. It is what it is; you need to be very patient. Having this mentality is a part of the necessity for being a good developer. I just mean that it's no surprise if implementation and raising a PR might be slow. The good news is I got a review for my last-second PR, and it sounds like the Besu team is softer on implementing Ephemery natively inside the codebase of Besu. I'm working on the review currently. The following is based on the roadmap for next two weeks. However My priority is working on the reviews on the last PR. > - Weeks 14 and 15 Research on the 'reset peers' task. Implement the task.