Trust

One of the many problems with UMN’s research is that it experimented on trust. UMN’s previous contributions to the kernel built some trust in their institution. Then they ran a fully UMN-sanctioned experiment on the trust-based patching process without disclosing they were UMN researchers. After the experiment was complete, they revealed their nature. Later, UMN researchers submitted another buggy kernel patch.

There’s no way for the kernel maintainers to know whether or not this buggy patch was submitted in good faith or if it’s part of the same experiment or if it’s some new line of research. There’s no way of telling if UMN’s responses to being banned from kernel contributions are in good faith or if the responses are more experiments. They didn’t disclose the research in the first place, so there’s no reason to assume they would say if this is a good faith response, nor should they be trusted if they did.

Remember, this wasn’t some wildcat research project. UMN knew what was being done. They risked their institution’s reputation on it, not just the reputations of the individual researchers.

By experimenting on trust in a process that relies heavily on trust, UMN have put themselves further back than a neutral no-trust reputation. UMN is in anti-trust. It should take a lot more than just submitting new patches in presumed good faith to rebuild that. This is the problem with purposefully deceiving people. When you get a reputation as a liar, you can’t just tell people that you’re telling the truth this time and expect them to believe you.

Background: https://arstechnica.com/gadgets/2021/04/linux-kernel-team-rejects-university-of-minnesota-researchers-apology/

https://cse.umn.edu/cs/linux-incident