(apologies to those who have heard about this from every librarian they know for the past month)
A major publisher announced at the end of April that they will seek to prevent green OA author's manuscripts* from appearing on publicly-accessible websites, including institutional repositories, within an embargo time of up to four years (depending on the journal). According to COAR (the Confederation of Open-Access Repositories), this is retroactive. They also announce a policy of noncommercial noderivs licenses on all copies of author's manuscripts, which would mean that we, for instance, couldn't use images from these sources, even after the embargo had expired.
*(post-peer-review, pre-publisher-copyediting-and-formatting, for complete clarity; see W:self-archiving)
This seems to have annoyed some repositories, and they are requesting signatures for their statement of opposition: https://www.coar-repositories.org/activities/advocacy-leadership/petition-ag.... The signatories make a nice list of possible institutional collaborators. It is possible the Wikimedia Foundation may also wish to comment.
The publisher does say that academics may publish their own articles on their personal blogs and websites (judging from past takedown requests, this appears not to cover personal webpages hosted by universities). Compared to institutional repositories, indexing and discovery would be more difficult and more fragile. So we have a substantial new use case for WikiData WikiProject Source MetaData.
Can we write something to post to the Confederation of Open-Access Repositories, essentially saying that we can help crowdsource links to historical papers, and maintain those links in future? How far do people think we are, practically, from being able to do this? Many of these repositories have good open APIs and good staff, so we could in principle upload a lot of metadata fast.
Regards, HLHJ