I give out every paper I've published for free. Well, actually every paper I've published is also available in essentially final form on the arxiv, so I'm not a

I give out every paper I've published for free. Well, actually every paper I've published is also available in essentially final form on the arxiv, so I'm not a necessary middleman. But I read a lot of papers from a lot of people that don't make their papers freely available and who don't use the arxiv (or any other preprint server). This seems particularly common for people a bit older than me. And of course there is the set of papers published more than 20 years ago, say, but which are still paywalled (or worse, essentially impossible to find anywhere). For these, scihub is great.