On Apr 29, 2024, at 5:15 PM, Fitchett, Deborah <[log in to unmask]> wrote:

> With open data sources, also emphasise checking for any throttling limits and if the site doesn't specify then go super slow: it may not be resourced for dealing with large number of queries and it's a bit rude to accidentally DoS attack an open site...


On Apr 30, 2024, at 4:38 AM, Owen Stephens <[log in to unmask]> wrote:

> Potentially you could advise them to take a look at CORE https://core.ac.uk/services which harvests from across many open repositories (metadata and full text) already and offers downloads and sync mechanisms. Downloads of older copies of the data are free and licenses ODB-BY. If they need the most recent data then that can be licensed for a fee, or accessed via a CORE “Sustaining” membership (if your institution has one, or is willing to take one - it does have some other benefits as well as access to the dataset so maybe worth looking at https://core.ac.uk/membership%23membership-levels


Again, plus-plus (++) for these ideas, and to my mind they describe directions our aquisisions departments ought to be going.

For example, a long time ago, libraries stopped cataloging journal articles because the process was too labor intensive. Instead, I believe, we subscribed to things like Reader's Guide to Periodical Literature. Such was functional and a good financial decision. Now-a-days, when we can harvest journal articles, complete with metadata (and/or programatically genearate our own metadata), it is possible to catalog journal articles again. OAI and Core are excellent starting places.

--
Eric Morgan
University of Notre Dame