Announcing Google Book Search, um, Search

From: Tim Spalding <tim_at_nyob>
Date: Mon, 17 Sep 2007 10:46:01 -0400
To: NGC4LIB_at_listserv.nd.edu
I've made a bookmarklet, called "Google Book Search Search," that
gathers Google Book Search IDs for titles in LibraryThing.

All data collected is available to you in its entirety. You can use it
to enhance your library catalogs, adding links to Google Book Search
for books that Google has. It also captures what they have, so you can
decide to link to them for books with full text, or whatever.

Here's the blog announcement, with a link to the bookmarklet:
http://tinyurl.com/34satk

NGC4Lib-relevant details:

* The Google Book Search ID is an impenetrable "hash" (for example,
L4q6Z0s_u14C). They have never provided a list of the books they have.
It's not clear to me if they are trying to hide what they have, or if
they just haven't prioritized this. It seems to me, however, that
libraries that join GBS in the future should at least take care to
require Google to make the hashes public. I at least am happy to
defend Google's right to hold on to its scans, but this is something
else OCLC asserts rights over bibliographic data, but it doesn't keep
the OCLC number secret.

* Others, like John Blyberg, have tried to assemble Google Book Search
IDs before, only to run into the Google-wide throttle on machine
requests. GBSS works because the action is distributed. Since
launching it at 2am, over *three hundred* LibraryThing users added the
bookmarklet and started scanning for books. If Google wants to shut it
down—which is their right—they're going to be clear about what they're
doing, restricting information about their holdings, so that people
and institutions cannot link to them reliably.

* Right now it looks for books in LibraryThing, either random ones or,
if you're a member, yours too. If libraries out there want their books
run through, we'd love to put them in. We're also willing to
open-source the code. The data is already free and available.

* Right now we're relating Google Book IDs to ISBNs, titles and
authors, as well as the LibraryThing "work" code. We need to dig LCCNs
and WorldCat numbers out of MARC records to make this as useful as it
could be.

--
Check out my library at http://www.librarything.com/profile/timspalding
Received on Mon Sep 17 2007 - 08:53:52 EDT