While it is tempting to agree, I feel that if you take these principles to their logical conclusions, you lose any incentive to actually participate in any open source projects. If I cannot expect the author to even look at my code or read my issue reports, there is no reason for me to write the reports or share my fixes. If all my dependencies are my responsibility, the logical way to go is to fork every code base I am going to use, review all the code, fix whatever issues I find in my fork and never share any of it with the original project because there is no reason to.