The first impression is, that the Physics and the computer science have much in common. A deeper look into the subject shows, that they are about completely different aspects of the world, and the inner working of the scientific community around these subjects is different. Physics can be called without any doubt a scientific discipline. It is teached at universities, has evolved over the year and the progress is done as a challenge between the best ideas. That means, if one physicians has a better idea, he has to prove that this idea works better then the known information before. At the end, all the physicians are unite for one idea and the opposition is called conspiracy theory or non-scientific claim.
Computer science works a bit different. The most remarkable difference is, that computerscience doesn't know any kind of opposition. There is no contrast between accepted theory and discarded theory, instead computer science works like a puzzle in which all the pieces are connected together. In the example of Linux vs. Windows, it is obvious that some kind of aggressive discussion is available, but the opposite direction is also part of computerscience. That means, there are no operating systems inside computer science which can be called unscientific. Every theory or sourcecode is welcome.
This working principle is uncommon for a scientific subject, it is more the work-mode in the human-sciences for example art, history and music. Music is perhaps a good example, in which also piece of music is called music. Even 12 tone music (which sounds very uncommon) is called official music, there are no works available which are non-music.
In a real science subject, it is very easy to contribute a non-sense argument. For example it is possible to invent a new kind of physics or a biology which is in opposition to main-stream biology. Under the term pseudo-science this is done very often and all theories are not backup-ed by academic science. The consequence are endless debates about what is true science and what not. In contrast in the field of computing it is nearly impossible to contradict the mainstream community. Even if somebody tries to brake the existing system and invents something which is complete non-sense his works is never out-of-the-box. Even joke programming language like Brainfuck or Joy are discussed seriously, that means, they are recognized as a correct contribution to the computing science.
Sometimes, computerscience was called an alchemy. This implies that in the future, it will become perhaps a science. I would argue, that computer science was never invented as science, it was invented as art. That means, it will stay on the same level like music or literature, that means it is not possible to decide which is part of computing and which is non-sense.
How could it be, that the obscure composition of Arnold Schönberg doesn't violate the music-science? Because music never was invented as a science. All musicians are composing their work from scratch, that means there is nothing like a mainstream theory about music. A piece of work can't be right, it has only an author and a date in which the work was published. In contrast, a real science like physics can be right or wrong. It is possible to teach the right physics or the wrong one, which isn't accepted by anyone. Identifying the opposite is a fundamental element of the university discussion. All the researchers are talking about the question if a theory is scientific or not. And if the researchers are great they are very conservative, which means they are calling nearly everything what is new non-scientific because it broke the rules. On a higher abstraction this is called convergent thinking. Scientific progress has to do with it, while arts works with divergent thinking which means to be open in any direction.