oiuyhtgyhuj
It's like the cable companies, they want your $100/month irrespective of whats showing - they don't want you to pay $1/show for what you actually want to watch.
Nature is paid to do two jobs.
1, Provide a copy of the paper on paper to sit in a library for 100s of years.
2, Certify that a paper has at least been written by somebody vaguely reputable and has been peer reviewed
Originally the first job was the main role, now with the internet the second is far more important. I am happy for my institution to pay for a paper copy of Nature in order that there is a universally recognised 'I am not a crank' stamp to put on my work. Think of it as an SSL-cert for my pages.
Yes I would like there to be a free open online web-of-trust for peer reviewed papers - but for now I have no objection to Nature doing that.
And I would much rather have a commercial publisher do it than have it mixed in with a professional body like IEEE/ACM/IOP who have a whole other set of agendas.
So I put a sample of code in a paper to explain an algorithm and I can no longer use that code, or include it in a GPLed work?
In fact between my university's policy that it owns anything I do that can be commercially exploited, the journal claiming it owns everything and I can't post it online and the various different international laws on software patents and publishing code that has anything to do with crypto/security - in theory I would spend a year talking to lawyers before publishing each paper.