I feel like a lot of this article is trying to make comparisons between online only open access journals and traditional closed publishers. However the paper th

I feel like a lot of this article is trying to make comparisons between online only open access journals and traditional closed publishers. However the paper the article is based on does not collect any data about the latter and so there isn’t any real comparison to make. I don’t think the solution is to move back towards the old model. There are already lots of initiatives towards creating online archives of academic work that may be piggybacked on. In mathematics, perhaps the easiest way to set up an open access journal is as an arxiv overlay journal where at the most basic level each issue of the journal is a list of links to specific versions of papers on the arxiv. This would be likely to be archived sufficiently well. For a traditional journal that shuts down to be archived, lots of things need to happen: 1. Some library needs to pay some exorbitant fee to get physical or (permanent not saas-based) digital copies of the journal 2. That library needs to keep hold of that copy for the 100 years or so until copyright expires 3. That library then needs to take the initiative to make its copies available This seems like a harder process than finding some public domain digital copy. And for a lot of journals, the only reason the library gets copies is due to the bundling systems which universities hate. I’m curious to know more about these journals which did vanish, and what sort of quality they are. If a predatory journal offers open access and later disappears, would they be counted?