Re: OCLC annoucement

From: Deborah Fritz <deborah_at_nyob>
Date: Thu, 23 Apr 2009 18:06:08 -0400
If you 'batchload' your records to OCLC to attach your OCLC code for those
records to matching records in WorldCat, OCLC is now offering to return your
records to you with the OCLC numbers of the matching WorldCat records
inserted in 035$a. The idea being that you will reload those records to your
database in order to get those OCLC numbers in the records in your database.
Hopefully your system is able to load only the 035 field rather than
overwriting your entire database record.

This certainly indicates that OCLC now believes that adding an OCLC number
to a record that is not really an OCLC record is OK. Which means that no one
can now trust that the presence of an OCLC number in a record means that the
record was ever 'an OCLC record'. 

So, if your librarys sends 'vendor' records to OCLC for batchloading to set
your holding code on WorldCat for those records, then you will get those
records back with OCLC numbers inserted and can load them back to your
database, set to match on record IDs, and you'll have OCLC numbers in those

Hmm, this announcement certainly goes a long way towards explaining why OCLC
is making this particular offer to batchloading libraries. 

Very interesting how this is all fitting together. 

My question is, how are libraries supposed to keep their local databases in
sync with WorldCat, for the 'Quick start' version of this; at least until
they replace their entire ILS with the OCLC Web ILS.


Deborah Fritz
MARC Database Consultant
The MARC of Quality
Voice/Fax: (321) 676-1904

> -----Original Message-----
> From: Next generation catalogs for libraries 
> [mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Jonathan Rochkind
> Sent: Thursday, April 23, 2009 5:38 PM
> Subject: Re: [NGC4LIB] OCLC annoucement
> Karen Coyle wrote:
> > What this essentially means is that if you have a Sirsi/Dynix, Ex 
> > Libris, etc., ILS, your current vendor has to serve up the 
> OCLC number 
> > consistently so that you can begin to replace them with OCLC. Hmmm.
> >   
> I don't think it's really a vendor-specific thing. If you 
> have an OCLC number in the fairly standard place in the MARC 
> record, I think any vendor's software will 'serve it up'.
> BUT. That IS a big 'if'.
> I'm not sure if even 50% of the records in my own catalog 
> actually have an OCLC number attached.  This will vary a lot 
> between libraries and their historical practice.
> But it's a non-trivial and difficult to automate process to 
> add OCLC numbers to the literally hundreds of thousands of 
> records that do not have them attached.
> One place where this is especially a problem is with records 
> purchased from a for-profit vendor -- an _increasingly_ 
> common phenomenon. These will almost NEVER have an OCLC 
> number attached, because the vendor does not provide one. In 
> fact, even if the vendor _wanted_ to do the extra work to 
> attach an OCLC number, they'd probably have to pay OCLC to do 
> so! Note to OCLC, you are working at cross-purposes.
> But there's an important distinction. When I bring this up to 
> catalogers, they often say "Well, those records CAN'T have an 
> OCLC number on them, becuase they are NOT OCLC records, it 
> would be a LIE."  
> Okay, they aren't OCLC records, but they represent the same 
> "manifestation" as a particular OCLC record. They _could_, at 
> least in theory,  have somewhere in the record an OCLC number 
> recorded NOT meaning "this is an OCLC record identified by 
> this #", but instead meaning "this represents the same 
> manifestation as the OCLC record with this #."  It would add 
> a LOT of value to libraries holding these records if they did 
> have such -- whether or NOT the library holding the record is 
> an OCLC member, in fact, there are plenty of things you could 
> do wtih this OCLC number.
> If OCLC is smart, and wants to succesfully make Worldcat and 
> Worldcat Local as ubiquitous as they're trying to -- they've 
> got to work with for-profit MARC services to get an oclcnum 
> on the (non-OCLC) record somewhere, indicating not that it's 
> an OCLC record, but that it represents the same manifestation 
> as that record. It's okay if some of them are wrong.
> Jonathan
> >   
> __________ Information from ESET NOD32 Antivirus, version of 
> virus signature database 4031 (20090423) __________
> The message was checked by ESET NOD32 Antivirus.

__________ Information from ESET NOD32 Antivirus, version of virus signature
database 4031 (20090423) __________

The message was checked by ESET NOD32 Antivirus.
Received on Thu Apr 23 2009 - 18:07:37 EDT