SimonForsman's stance on public universities is ethically correct. Public universities (all materials and employees, including e.g. professors and research assistants) in the EU are paid exclusively by tax funds. Insofar it is only just and reasonable that what they produce belongs to the public. Of course, students aren't being paid, so that's a different thing.

I might be wrong about this, as I am not an expert on the EU university system, but I find it very improbable that tax payers fully support university research. They might support a lot of it, but it's pretty common for private industry to treat university labs as cheap ways to incubate new and risky ideas, even if it's just to groom current PhD's for future employment in their own company. A lot of this comes in the form of joint efforts or industry grants, or even things as simple as material donations (like computers), all the way up to joint labs where a company might have a few full time workers literally physically in the university lab doing research or guiding projects. This is especially the case in engineering feilds, like CS.

I know my lab had recieved upwards of 300K in hardware, along with some one-of-a-kind prototype stuff, from industry, along with hundreds of thousands of dollars in funding that went to pay PhD student's living expenses and tuitions. There was also a bunch from governmental sources as well.

Industry has to have some expectation of transparency when working with university programs, as grad students need to publish in order to develop themselves. I can tell you for sure though that our industry partners would have been a lot more nervous about sharing their internal secrets with us if they thought we were required to release source material. We kept secrets when it was necessary to keep our relationship intact.

As a small side-note, one of the worst decisions our group ever made was releasing source for an internal project. Not only did it absolutely flood us with requests from other groups/students to provide support to whatever question they had, but we found ourselves racing against other groups to cover topics that we had already talked about, but hadn't explored yet. Having some other group slide under you and publish what you were trying to publish, and use your own infrastructure to do it, totally sucks.

From a users perspective too, there is a certain advantage to limiting released projects to only those that are 'ready for prime-time'. If your work is a hacky mess, you only would be reducing the signal/noisy ratio by releasing it. Leaving that as a decision to be made by the group lets them do it at their own schedule.