On 21 July 2011 17:52, Gwern Branwen <gwern0(a)gmail.com> wrote: The watermark is in a separate PDF layer, so is trivially removable before upload. I have just suggested on internal-l that someone give JSTOR a call, giving them the chance to look good by releasing this stuff themselves. Given that US law is utterly and unambiguously against them having *any* control via copyright over this stuff, no matter what silly deals they may or may not have signed with the Royal Society. (I'd be uploading them myself now except I'm in the UK, where the copyright situation surrounding slavish copies of PD works is still in a state of quantum uncertainty.) - d.

On 21 July 2011 20:28, Stephen Bain <stephen.bain(a)gmail.com> wrote: Indeed. But everyone would be happier if JSTOR stopped trying to enclose the public domain. It would be nice for various purposes if this question didn't come up, but events are well ahead of any such blissful state of undecidedness. I expect quite a few uploads of public domain content from JSTOR in reasonably short order. The only thing is that every uploader needs to check every scan for JSTOR additions or manipulations that may be intended to claim a new copyright. (Though this would arguably involve misstatement on JSTOR's part that the document was actually what they claimed it was.) - d.

On 7/21/11 10:59 PM, Andrew Gray wrote: I agree that JSTOR aren't the only ones at fault, but my interactions with them have generally not given the impression that they're interested in serving the public domain, either. If anything, they have a very protective attitude towards "their" database, taking a more aggressive stance than many of the journals who actually own the content in question. They've been completely uninterested in developing any sort of free-access policies, despite the fact that, from what I know from at least one journal, some journals would in fact make some of their old content freely available through JSTOR, if JSTOR offered that as an option that they could choose--- something like the NYTimes, "pre-1923 free, post-1923 pay" archive policy. But JSTOR doesn't even allow a journal to mark any portion of their archive non-paywalled, much less actually push for anything like that. In fact, they even bargain fairly stingily when it comes to temporary and partially free access. For example, the "19th Century British Pamphlets" collection was scanned thanks to a public research grant, which as a condition required JSTOR to give free access to all UK educational institutions through 2019. But they wouldn't agree to make it completely free, or to offer free access for more than 10 years; that's JSTOR's unwillingness to let go of control over their archive, the funding body did not demand a 2019 sunset for free access, or the restrictions on who could access it. I hope you're right that they can be encouraged to engage in more productive collaboration in the future, but for my part I'm hugely disappointed and disillusioned with them. At one point in the late-90s it seemed like they might become something of a larger-scale Perseus Project, balancing a need for continued funding of their project with a mission of digitizing humanity's common heritage and making it freely available online. But I haven't seen any evidence of their leadership having that kind of vision. -Mark

On 22 July 2011 02:21, Gwern Branwen <gwern0(a)gmail.com> wrote: I've just been doing the same figures, albeit focusing on 2008: http://www.generalist.org.uk/blog/2011/jstor-where-does-your-money-go/ In short: - the "joining fee" is a bit over the ongoing journal scanning costs, but not dramatically so - the annual fee then gets split 30% publishers, 70% JSTOR JSTOR merged with another body in 2009, so the 2009 figures include some of its overhead and operations - the JSTOR-project-specific staff cost, for example, seems to be a good bit lower judging by 2007/8 data. I've used 2008 data to try and correct for this. The one detail that really leapt out at me was how few once-off individual users there are. Total income from "pay per view articles" in 2008 was under $150k - perhaps representing seven or eight thousand articles - and dropped to half that in 2009. Given how ubiquitous JSTOR access is at academic institutions, it makes sense that it would be rarely used, but still. It didn't appear the year before, and I suspect may be something to do with the merger & corresponding transfer of assets. My interpretation here is that the IT costs may be purely "physical" and not include the development staff - the depreciation figures seem too high to make sense otherwise. There's no staff breakdown, though. -- - Andrew Gray   andrew.gray(a)dunelm.org.uk