3 ms·
All JSTOR did was try repeatedly for several months to block massive downloads that were causing service disruptions. JSTOR never involved law enforcement in th
by tzs 6y ago
All JSTOR did was try repeatedly for several months to block massive downloads that were causing service disruptions. JSTOR never involved law enforcement in this, and even tried to contact the downloader to see about arranging less disruptive access. Instead, the downloader kept evading the blocks.
It wasn't until MIT discovered that the downloader had moved from using their public guest WiFi network to going into places closed to the public to surreptitiously install hardware directly on the MIT LAN to continue the downloading that MIT called law enforcement.
Here's the timeline from JSTOR's point of view [1]. What do you see in there that they should have done differently?
Keep in mind that JSTOR is essentially a library and a non-profit. They do not own the copyright on the journals they distribute. Wherever they can, they provide cheap or free public access to most of the library.
They have probably done more than any other organization to provide ways for the public to get free or cheap access to non-open access journals. Maybe someday most research will be published in open access journals and we won't need JSTOR, but until then they are among the good guys.
[1] https://docs.jstor.org/summary.html https://docs.jstor.org/summary.html
- jancsika 6y ago> We frequently support such research by providing access to datasets, free of cost, in a way that does not affect access for other users. This is a bit misleading. They are talking about large datasets here. But in a later paragraph they describe their central fear-- that the downloader was trying to acquire the entire data set. And-- according to JSTOR-- while they feared the entire dataset was being downloaded, the access pattern had changed so that it was no longer interrupting regular JSTOR access at MIT. So disruption to service was clearly not the issue at that point. It seems reasonable to conclude from this, and from their subsequent actions, that their definition for "large datasets" does not include "the entire dataset." As the documentary points out, there was precedent for Swartz downloading entire datasets from other databases for legitimate research purposes prior to this event. Is there evidence that JSTOR allowed other researchers to run automated analyses on the entire dataset through their free service? If not, they should have stated explicitly somewhere in their statement that they believe it is not legitimate research to run analyses on their entire dataset.
- blindm 6y ago> going into places closed to the public to surreptitiously install hardware directly on the MIT LAN I think that's a line few hackers cross. I mean physically breaking and entering a server room to help with exfiltration all whilst being monitored by CCTV is pretty bold stuff.
- Zigurd 6y agoIt was a wiring closet. I've been in some of these, possibly the same one. While there is more valuable stuff than telephone punch-down blocks in them nowadays, it is nothing like breaking in to a server room.