Nature News is a weird place. They are editorially independent from main Nature, and as such they often make the 'right noises': articles decrying the ever-lasting hunt for impact factors and glam journals, articles hyping up open access; all the while the main Nature only exists, and makes a lot of money, because it is so central to bad publishing practices.
Paywalls are still common. But mostly with legacy publishers that are increasingly marginalized by newer ones. When it comes to scientific publications there are a few holdouts. But by and large, scientists and universities are increasingly reluctant to spend significantly on paywalls.
It used to be more or less automatic that universities would allocate huge chunks of public funding to maintaining all sorts of subscriptions. Tens of thousands of our EU project budgets were actually earmarked for this in the nineties. There was no debate about any of that at the time. These days there is a bit more scrutiny and quite a few universities are refusing to do this at this point.
I actually had to go to a library in the nineties to get my hands on some articles when I was doing my PhD. But even then, this was rare for me. So rare in fact that I remember these handful of separate occasions where I had to do this. Even before Google launched, you could find a lot of .ps and .pdf files online. And in the rare case I couldn't, I often opted to ignore the work. I was skimming through papers by the dozens every week. Google made this stuff a lot easier when it launched. Computer scientists were early adopters of paywall evasion of course.
Something I figured out early was that as a scientist, making life easier for other scientists to access my work is not optional. The whole point of publishing your work is getting other people to read it. The best way of ensuring that never happens is paywalls. So, I put my articles on my website in the hope that people would actually read them.
The reason scientists continue to publish to pay-walled publications is that they are financially incentivized by impact metrics. You publish in something like nature, it looks great and your metrics improve and your boss loves you. Of course, most scientists don't actually get published there and instead have to deal with some severe yawn inducing second or third rate journals by the likes of Elsevier, Springer, and others. Been there done that.
And since it's all about impact, getting published is only half of the success. The other half involves getting people to reference your work. Which requires them to be able to access it. Which pretty much requires you to enable that by putting the pdf somewhere for people to download. Either you do that or nobody reads your papers.
Hence the practice of these articles in paywalled publications getting published online elsewhere. Paywalls have never been good at preventing that or putting a stop to that practice. Scientists actively depend on enabling paywall evasion to get their metrics up.
> Hence the practice of these articles in paywalled publications getting published online elsewhere.
OK, if you know a set of these elsewheres, perhaps you can help with this.
What's the absolute minimum viable research subscription (open first, then paywall) sites required to search and get access to articles for by-and-large all* research papers in science, technology, engineering, and math?
What about even just technology?
As a private individual who reads a lot of research, over past decade, I've tried several times to figure out the minimum viable total subscription coverage by starting here:
