Skilled students can make some very fine PRs when they are working on something that intrigues them enough that they get very immersed in the code. But not all

Skilled students can make some very fine PRs when they are working on something that intrigues them enough that they get very immersed in the code. But not all students in training are skilled, and making code contribution a required part of a program seems certain to reduce the average quality of PRs. The problem is that somebody has to review those PRs, and often interact with the proposer for quite a while, to teach them the approach that is desired for the individual project. PRs that go against the grain are not always worth the effort. In the scenario of the article, open-source developers are saddled with a new task: teaching students, many of whom will be unmotivated by anything more than a grade. No thanks. Hard pass. If professors want students to learn how to interact in a group setting, then the students ought to be designing their own software. If there's a perceived need for larger-scale projects, the professors ought to set them up and maintain cohesion and momentum across terms. If the professors don't have ideas for such projects or the time to spearhead them, well, then that's a bit of a sign.