Hi, What lively discussion!
> Which begs the question, what /is/ the mission and goal of the OL project? We are building an open, editable database of books. You've all read the tag lines like "A page on the web for every book," and "a Wikipedia for books" - these are other descriptors of the project. > So far, there's been a lot of tap dancing around the question of whether > OL can or will support MARC exports, MARC-XML exports, or Z39.5 access, > but no one with presumptive authority to speak on behalf of Open Library > has come forward to say "we can," "we can't," "we will," or "we won't." We looked into this, and discovered that it's possible to get Z39.50 access to Open Library records via biblios.net: http://www.liblime.com/documentation/faq/faqsection_view?section=Cataloging (scroll to the bottom) If anyone is interested to have a look around in there, please let us know if that is adequate? (Open Library doesn't interact directly with any Z39.50 services directly.) > Reading between the lines of Ms. Coyle's posts, I conclude that the > answer is "we can't, therefore we won't." Had the response been framed > this way, it would have avoided a fair amount of discussion. The discussion has been useful for me. > So I ask again, who are the intended consumers of OpenLibrary catalog > data?, what is the intended use for OpenLibrary catalog data?, and how > well does the current interface address the documented intended use? Even back as far as November 2008, it was Aaron Swartz who said "Our primary audience isn't librarians, it's web users and developers. We're not out to replace OCLC. We decided that we should import the data that's most important to them and keep links back to the original MARC for the librarians." We are doing this. In the redesign, we took away 2 clicks to view the original source MARC (or link to original record on Amazon). Aaron also mentioned that if this was something libraries would be prepared to pay for, we could work on making it easier to extract original MARC records in bulk, but that is not our primary goal. As I mentioned, the primary goal is to build a comprehensive database of books. > Another acceptable response might have been, "we don't but we're working > on it." Or, "we don't but we will if we can figure out how." Our answer is generally "we don't, but we will if we can figure out how" :) It's a concern (to me) that people perceive we aren't trying to have records flowing freely into and out of Open Library. We've had an open (and used) API for several years now. I must say also that the perception that every MARC record we receive is excellent and pristine is just wrong. As Karen mentioned, there are a lot of partially complete and just plain wrong records. We've begun several recent initiatives in the data exchange area: * Mass-write of Good Reads IDs to Open Library records (over 3 million updates) - We've received a data file from Library Thing too, for the same purpose - thanks Tim - we'll get to it as soon as the dust from our recent full redesign settles. * We're also working internally here at the Archive to see how our database of books can inform and support the scanning workflow. It seems like a no-brainer that the Archive should use OL data to initiate the scanning process ("find a record for the book I'm about to scan"). The thought here is that whatever processes we put in place to support this workflow could be used externally. We're looking at an atom representation of OL records, and intend to match DC standards there wherever possible. (We've literally just started on the details of that, this week.) * We have built an alpha version of a tool to help users merge duplicate author records. We're still testing it, and this needs to be both stable and undo-able, so it will be some time before it is released to all, and even then will probably only be possible by signed in users. I also suspect it might be only a handful of libraries out there who would be willing to import MARC edited by the general public back into their catalog. Anyone know any different? (I have personally made several hundred straightforward corrections to records. Nothing fancy, just fixing typos or wrong publish dates etc.) So, a question back to the group is: do you think you and your teams would be able to consume an atom feed? > it may be because development resources are scarce It's probably stating the obvious, but we're understaffed and hiring and Open Library is a fun project! Also, please note that Open Library is a free service, funded entirely in 2010 by the Kahle/Austin Foundation. Cheers, george _______________________________________________ Ol-tech mailing list [email protected] http://mail.archive.org/cgi-bin/mailman/listinfo/ol-tech To unsubscribe from this mailing list, send email to [email protected]
