13 Dec 03:12
Re: Laundry list for NGC (long post) -- How does Koha measure up?
From: Tim Spalding <tim <at> LIBRARYTHING.COM>
Subject: Re: Laundry list for NGC (long post) -- How does Koha measure up?
Newsgroups: gmane.culture.libraries.ngc4lib
Date: 2006-12-13 02:12:36 GMT
Subject: Re: Laundry list for NGC (long post) -- How does Koha measure up?
Newsgroups: gmane.culture.libraries.ngc4lib
Date: 2006-12-13 02:12:36 GMT
Karen (et al.), So, I'm hasty. Let's do it. I really want to help and indeed for LibraryThing to take the lead on this—with everything being fully downloadable and open source, of course, so LibraryThing doesn't have any real power over it. I whacked together the first LibraryThing in a month. This is a weekend job, I think. Caveat: The very real beauty of Ruby notwithstanding, I'm not going to help unless it's in PHP. (I'm limited, and PHP is the most widely shared hacker language.) Some initial thoughts/responses: 1. Code should be CC. 2. There's a lot of PHP code from LibraryThing we'd be glad to release in the CC. Image downsampling, ISBN checking, etc. 3. Covers can't actually be CC. In my understanding, covers retain their copyright, it's just that showing them in our contexts is automatic fair use. If every bookstore needed to get permission to show a book cover—or a poster or etc.—in an advertisement, chaos would ensue! That's how covers work. But you can't take a piece of a cover and use it for some other purpose. For starters, many covers show copyrighted, permissioned art. 4. The data should be freely available for putting on other servers, but if LibraryThing hosts the first copy, we obviously reserve the right not to provide ten terabytes/day to one IP. We just paid for another tube, though, so we've got bandwidth to spare. It's not like we're talking video here. 5. I'd love to hear what you think, but I propose going as minimal as possible with the bibliographic data. The primary access should be by a URL, e.g., http://www.coverthing.com/small/006076550X (yes, I just bought that domain, and bookdatathing.com for good measure <grin>). That would return either a cover or a 1x1 transparent pixel, just like Amazon. Consequently, if there's an ISBN, it should be by ISBN (10 and 13 handled equally). If not, I think it's by LCCN, if there is one. If not, we can come up with a unique key and provide a way to search for it, but its unlikely to be found. 6. Starter interface is a simple form—as simple as possible. ISBN, type (cover is default) and the upload itself. It should be easy as pie to upload images. There are, for example, thousands of Amazon users who gave Amazon their "customer images" just because it felt good—many have given hundreds of them. And LibraryThing is *constantly* sent cover art by authors, who don't know that we can't used it. 7. I propose we adopt some standard sizes. They would be downsampled from what you start from, which would always be retained. 8. I'd also like to push for some optional variants, like (a) cover image (b) back cover (c) spine image (d) other user image. LibraryThing users want spines, I know. 9. I can contribute a mess of data. In theory, LibraryThing can contribute its 230,213 (!!) user-provided covers. (I need to read up more on some legal issues involved before I stake the company on it.) And, although I'd like to check with the publishers just to make sure, we can contribute the ONIX feed covers too. (I don't know the number, but it's big, albeit focused on two publishers). Okay, breathe, breathe. ---------- Forwarded message ---------- From: Karen Coyle <kcoyle <at> kcoyle.net> Date: Dec 12, 2006 8:14 PM Subject: Re: Laundry list for NGC (long post) -- How does Koha measure up? To: NGC4LIB <at> listserv.nd.edu Chris Cormack wrote: > That sounds like a great project to get content and put it under a > creative commons type license. With the buy in from the publishers of > course. > And if we did it, such that it was general enough (the application > that is not the content) such that it could be used by others for > other purposes. It would perhaps be a useful Open Source project. > > Yes, great idea! So, what would be the best link between the library catalogs and the covers? Is the ISBN enough? If we don't have a paperback cover, could we link to the cover for the hardback? etc. I'm game to throw in my bib-metadata knowledge if needed. We would have to get the publishers to agree on a CC license, and we'd need to get the license image or link embedded in the cover image or linked to it in some way. OK, I'm jumping ahead myself now. Need to go take a few deep breaths. kc -- ----------------------------------- Karen Coyle / Digital Library Consultant kcoyle <at> kcoyle.net http://www.kcoyle.net ph.: 510-540-7596 fx.: 510-848-3913 mo.: 510-435-8234 ------------------------------------
RSS Feed