A fascinating little Open Access kerfluffle with ACS Publications

I found this short piece from C&EN's Sophie Rovner pretty interesting:

Online Access To ACS Publications Is Restored After Some Customers Were Unintentionally Blocked 

On Wednesday, a blog published information about a security feature that the American Chemical Society uses to protect the stability of its website and prevent unlicensed wide-scale downloading of its content. Curious readers who then clicked on the link provided in the post ended up disabling their institutions’ access to ACS publications, which include more than 40 journals as well as C&EN. The society has already restored access for the sites of approximately 200 affected customers. 

Known as a “spider trap,” the security feature is designed to be triggered not by individual users but by automated website crawling and data extraction. 

The spider trap information was posted by University of Cambridge chemist Peter Murray-Rust after he was told about it by an unnamed source, whom he dubbed Pandora. Murray-Rust reported that Pandora found the link associated with an online ACS journal article and clicked on it thinking it would lead her to another journal paper. She then received an automated message saying her institution’s access to ACS publications had been blocked. 

Those who have clicked on the link and have unresolved issues can e-mail support@services.acs.org with their institution name, and ACS will work to reinstate access as quickly as possible.

The original blog post by Peter Murray-Rust is pretty interesting -- read the comments to see all the people who were clicking the Spider Trap Link of Doom, and then finding out that they had accidentally cut off their institution from access. (Onion headline: "Fog in Channel, ACS cut off")

I don't claim to understand all the software issues going on, but it is indeed interesting that ACS Publications is deploying these sorts of security features and willing to cut off access to an institution's IP addresses to make a point. I would love to know if these sorts of features are special to ACS or if other publishers also use these sorts of traps. From Dr. Murray-Rust's second post, it seems like they're more common than we knew...