3 ms·
There seems to be a lot of conflict of interests here. For one thing you could reject the code during code review then use it anyways, basically plagiarism. The
by SteveNuts 6y ago
There seems to be a lot of conflict of interests here. For one thing you could reject the code during code review then use it anyways, basically plagiarism. The other issue I could see would be people creating bugs on purpose to squash them, lowering the overall code quality and creating churn.
- jimmy2020 6y agoThe whole idea is based on goodwill. Why anyone rejects the code then use it anyway. The second issue is valid for open and closed source. Free or paid. There are always people who want to contribute regardless of the value they bring, and that doesn't affect the quality of the code because you aren't merging without reviewing the changes.
- btilly 6y agoThe biggest problem is people submitting code, it being rejected, a working solution using similar principles being accepted, and then accusations of plagiarism. With the same people always submitting the accepted code. The fact is that their work gets accepted because they followed coding standards and understand how to not undermine the project design. But people who have not internalized coding standards or the overall project design will not see that and get upset that they don't get paid for at least showing how to solve the problem. (Nevermind that the part of their solutions that they point to would have been obvious to anyone.)
- jobigoud 6y ago> The other issue I could see would be people creating bugs on purpose As a side note this particular perverse incentive also exists in the world of proprietary code security bug bounties. That's why bounties cannot grow to truly enormous amounts.