Re: Spiderable OPACs and the elephant in the library lobby

From: Casey Durfee <Casey.Durfee_at_nyob>
Date: Wed, 25 Apr 2007 10:23:57 -0700
To: NGC4LIB_at_listserv.nd.edu
 
I don't think there's really a resource issue at all.  Say there are 10,000 libraries in the world and they each own 1,000,000 titles. That's only 10 billion datapoints -- probably less than 100 GB of data total.  You could run such an operation on a $20/month shared web hosting account (1and1 and Dreamhost both offer around 300 GB of storage and 3 TB of bandwidth a month for that price -- that's more than enough for a "little O", I'm willing to bet).  Nor is it a huge technical issue -- most ILSes make it possible one way or another to get your holdings data out.  I think the *4Lib crowd could crank out scripts for most of the major ILSes in a day or two.  
 
The huge issue is, how do you get every single library in the world to send you their holdings data?  

>>> William Denton <wtd_at_POBOX.COM> 4/24/2007 6:35 PM >>>
On 24 April 2007, K.G. Schneider wrote:

> I wonder how many of us are conceptually on board with the concept of a
> national catalog, and yet hesitate to endorse this concept (or even
> argue for a functional model we realize is not working for us now, if it
> ever did) because the only functional model remotely available to us
> (and not that remote any more, either) would place us under the control
> of the Big O.

National catalogues mean lots of catalogues!  There are lots of nations.

I met Thom Hickey, Chief Scientist at OCLC, a couple of years, and have
never forgotten that he said he'd take the WorldCat database home on his
iPod.  WorldCat's grown a lot since then, but so has iPod storage.

So disk space isn't a problem, but running something as big as WorldCat
does require a lot of resources.  On the other hand, it seems to be mostly
a matter of implementation.  What a large union catalogue does is mostly a
solved problem.

Is it the sort of problem where someone outside, like a Google or an
Internet Archive, would look at the requirements and say, "That's not that
much, actually--we have the disk space and bandwidth, and it's not much
work to build some basic tools and APIs"?

Maybe someone will treat building a freely available global union
catalogue as a corollary to their main business.

Bill
--
William Denton, Toronto : www.miskatonic.org www.frbr.org www.openfrbr.org
Received on Wed Apr 25 2007 - 11:20:59 EDT